International Energy Agency Technology Roadmap H2 FuelCells

2050
Hydrogen and Fuel Cells
Secure
Sustainable
Together
tiv
ec
Te
Technology Roadmap
rg
2040
Ene
2035
es
2045
chn
rs
olo g y P e
INTERNATIONAL ENERGY AGENCY

The International Energy Agency (IEA), an autonomous agency, was established in November 1974.
Its primary mandate was and is two-fold: to promote energy security amongst its member
countries through collective response to physical disruptions in oil supply, and provide authoritative
research and analysis on ways to ensure reliable, affordable and clean energy for its 29 member
countries and beyond. The IEA carries out a comprehensive programme of energy co-operation among
its member countries, each of which is obliged to hold oil stocks equivalent to 90 days of its net imports.
The Agencys aims include the following objectives:
n Secure member countries access to reliable and ample supplies of all forms of energy; in particular,
through maintaining effective emergency response capabilities in case of oil supply disruptions.
n Promote sustainable energy policies that spur economic growth and environmental protection
in a global context particularly in terms of reducing greenhouse-gas emissions that contribute
to climate change.
n Improve transparency of international markets through collection and analysis of
energy data.
n Support global collaboration on energy technology to secure future energy supplies
and mitigate their environmental impact, including through improved energy
efficiency and development and deployment of low-carbon technologies.
n Find solutions to global energy challenges through engagement and
dialogue with non-member countries, industry, international
organisations and other stakeholders.
IEA member countries:

Australia
Austria
Belgium
Canada
Czech Republic
Denmark
Estonia
Finland
France
Germany
Greece
Hungary
Ireland
Italy
Japan
Korea
Luxembourg
Netherlands
New Zealand
Norway
Poland
Portugal
Slovak Republic
OECD/IEA, 2015
Spain
International Energy Agency
Sweden
9 rue de la Fdration
Switzerland
75739 Paris Cedex 15, France
Turkey
www.iea.org
United Kingdom
United States
Please note that this publication
is subject to specific restrictions
that limit its use and distribution.
The terms and conditions are
available online at www.iea.org/t&c/
The European Commission

also participates in
the work of the IEA.
Foreword
Current trends in energy supply and use
are patently unsustainable economically,
environmentally and socially. Without decisive
action, energy-related emissions of carbon dioxide
(CO2) will more than double by 2050 and increased
fossil energy demand will heighten concerns over
the security of supplies. We can and must change
our current path. However, this will take an energy
revolution and low-carbon energy technologies
will have a crucial role to play. Energy efficiency,
sources of renewable energy, carbon capture and
storage (CCS), nuclear power and new transport
technologies will all require widespread deployment
if we are to achieve reductions in greenhouse gas
(GHG) emissions. Every major country and sector of
the economy must be involved. The task is urgent
if we are to make sure that investment decisions
taken now do not saddle us with suboptimal
technologies in the long term.
Awareness is growing of the need to turn political
statements and analytical work into concrete action.
To drive this forward, in 2008 the G8 requested
the International Energy Agency (IEA) to lead the
development of a series of roadmaps for some of
the most important technologies. By identifying
the steps needed to accelerate the implementation
of radical technology changes, these roadmaps
will enable governments, industry and financial
partners to make the right choices. This will, in turn,
help societies make the right decisions.
Hydrogen and fuel cell technologies, once they
are more developed can support climate change
and energy security goals in several sectors of the
energy system, such as the transport, industry,
buidings and the power sector. Hydrogen can
connect different energy sectors and energy
transmission and distribution (T&D) networks,
and thus increase the operational flexibility of
future low-carbon energy systems. It can help to:
1) achieve very low-carbon individual motorised
transport; 2)integrate very high shares of variable
renewable energy (VRE) into the energy system;
3)contribute to the decarbonisation of the industry
and the buildings sector.
Although the GHG mitigation potential of hydrogen

technologies is promising, important obstacles for
widespread deployment of hydrogen and fuel cell
technologies need to be overcome. These barriers
are mainly related to current costs of fuel cells and
electrolysers, the development of a hydrogen T&D
and retail network, as well as the cost efficient
generation of hydrogen with a low-carbon footprint.
Most hydrogen and fuel cell technologies are still in
the early stages of commercialisation and currently
struggle to compete with alternative technologies,
including other low-carbon options, due to high
costs. Additional attention will be required before
their potential can be fully realised. Governments
can help accelerate the development and
deployment of hydrogen and fuel cell technologies
by ensuring continued research, development
and demonstration (RD&D) funding for hydrogen
generation and conversion technologies, such as
electrolysers and fuel cells. This will facilitate early
commercialisation of fuel cell electric vehicles and
support demonstration projects for VRE integration
using hydrogen-based energy storage applications.
Overcoming risks related to investment in
infrastructure hinges upon close collaboration
among many stakeholders, such as the oil and gas
industry, utilities and power grid providers, car
manufacturers, and local, regional and national
authorities.
This publication is produced under my authority as
Executive Director of the IEA.
Maria van der Hoeven

Executive Director
This publication reflects the views of the International Energy Agency (IEA) Secretariat but does not necessarily reflect
those of individual IEA member countries. The IEA makes no representation or warranty, express or implied, in respect
to the publications contents (including its completeness or accuracy) and shall not be responsible for any use of, or
reliance on, the publication.
Foreword
Table of contents
Foreword
Table of contents
Acknowledgements
Key findings
Cross-cutting opportunities offered by hydrogen and fuel cells
Energy storage and utilisation in transport, industry and buildings
Key actions in the next ten years
Cross-cutting opportunities offered by hydrogen and fuel cells
Energy storage and utilisation in transport, industry and buildings
Introduction
Rationale for hydrogen and fuel cell technologies
Purpose, process and structure of the roadmap
11
Roadmap scope
11
Technology status today
12
Hydrogen in transport
12
Hydrogen for VRE integration
19
Hydrogen in industry
24
Fuel cell technology in buildings
25
Other niche applications based on fuel cell technologies
27
Key hydrogen generation technologies
28
Key hydrogen conversion and storage technologies
30
Vision for deployment to 2050
34
Transport
34
VRE integration
47
Industry
53
Synergies between energy sectors
54
Parameters of key technologies today and in the future as used in the model
56
Hydrogen technology development: Actions and milestones
58
Data assessment and model development
58
Technology development
59
Policy, regulatory framework and finance: Actions and milestones
64
65
Hydrogen in stationary applications
66
The role of codes and standards
67
Finance
68
International collaboration
69
Social acceptance and safety
69
Technology Roadmap Hydrogen and Fuel Cells
Conclusion: Near-term actions for stakeholders
70
Abbreviations, acronyms and units of measurement
72
References
73
List of figures
Figure 1. Energy system today and in the future
10
Figure 2. Well-to-wheel (WTW) emissions vs. vehicle range for several technology options
14
Figure 3. Cumulative cash flow curve of hydrogen stations in the early market phase
17
Figure 4. T
odays carbon footprint for various hydrogen pathways and for gasoline
and compressed natural gas in the European Union
18
Figure 5. Electricity storage applications and technologies
20
Figure 6. Current conversion efficiencies of various hydrogen-based VRE integration pathways
21
Figure 7. Limitations on the blend share of hydrogen by application
23
Figure 8. Ene-Farm fuel cell micro co-generation cumulative sales, subsidies and estimated prices, 2009-14 27
Figure 9. Schematic representation of technology development potential of different electrolysers
29
Figure 10. Production volumes of fuel cells according to application
30
Figure 11. Production cost for PEMFCs for FCEVs as a function of annual production
31
Figure 12. Energy-related carbon emission reductions by sector in the ETP 2DS
34
Figure 13. PLDV stock by technology for the United States, EU 4 and Japan in the 2DS high H2
36
Figure 14. Cost of hydrogen as a function of electricity price and annual load factor
37
Figure 15. S
pecific PLDV stock on-road WTW emissions by technology for the United States,
EU 4 and Japan in the 2DS high H2
39
Figure 16. Scheme of hydrogen T&D and retail infrastructure as represented within the model
40
Figure 17. Hydrogen generation by technology for the 2DS high H2 in the United States, EU 4 and Japan
42
Figure 18. Hydrogen production costs without T&D for the 2DS high H2
42
Figure 19. Hydrogen stations for the 2DS high H2 in the United States, EU 4 and Japan
43
Figure 20. Vehicle costs, fuel costs and TCD for FCEVs in the 2DS high H2 in the United States
45
Figure 21. Subsidy per FCEV and share of annual subsidy as a percentage of petroleum
fuel tax revenue under the 2DS high H2 in the United States, EU 4 and Japan
46
Figure 22. C
O2 mitigation from FCEVs in transport under the 2DS high H2 in the United States,
EU 4 and Japan
47
Figure 23. Global electricity generation mix under the 6DS and 2DS
48
Figure 24. Installed electricity storage capacity for selected regions today and in 2050 under the 2DS
and the storage breakthrough scenario
48
Figure 25. L COE for inter-seasonal energy storage via power-to-power systems and VRE integration
via power-to-gas systems in 2030 and 2050
51
Figure 26. LCOE of different energy storage technologies for daily arbitrage in 2030 and 2050
52
Figure 27. M
arginal abatement costs of different hydrogen-based VRE power integration
applications in 2030 and 2050
52
Figure 28. Electricity price arbitrage and hydrogen generation costs
55
Table of contents
List of tables
Table 1. Workshops parallel to the development of the Technology Roadmap on Hydrogen and Fuel Cells
11
Table 2. Current performance of hydrogen systems in the transport sector
13
Table 3. Existing FCEV fleet and targets announced by hydrogen initiatives
13
Table 4. Qualitative overview of hydrogen T&D technologies for hydrogen delivery in the transport sector
16
Table 5. Existing public hydrogen refuelling stations and targets announced by hydrogen initiatives
17
Table 6. Current performance of hydrogen systems for large-scale energy storage
19
Table 7. Qualitative overview of characteristics of geological formations suitable for hydrogen storage
22
Table 8. Current performance of fuel cell systems in the buildings sector
26
Table 9. Current performance of key hydrogen generation technologies
28
Table 10. Current performance of key hydrogen conversion, T&D and storage technologies
32
Table 11. Cost of PLDVs by technology as computed in the model for the United States
38
Table 12. Techno-economic parameters of FCEVs as computed in the model for the United States
38
Table 13. Power-to-power and power-to-gas systems included in the analysis
49
Table 14. System specifications for inter-seasonal energy storage and arbitrage
49
Table 15. P
arameters used in the model for stationary hydrogen generation and conversion
technologies as well as for energy storage and VRE integration systems today and in the future
56
Table 16. Initiatives and public-private partnerships to promote hydrogen and fuel cell technologies
64
List of boxes
Box 1. Risks associated with investment in hydrogen refuelling stations
16
Box 2. Carbon footprint of hydrogen in transport
18
Box 3. Power-to-gas in Europe: storage potential and limitations
22
Box 4. The Japanese Ene-Farm experience
26
Box 5. ETP scenarios and the hydrogen roadmap variant
35
Box 6. The economics of renewable hydrogen
37
Box 7. Spotlight on hydrogen generation
41
Box 8. Vehicle costs, fuel cost and TCD
44
Box 9. Electrolysers in the control power market
54
Acknowledgements
The IEA Energy Technology Policy Division
prepared this publication. AlexanderKrner was
the project leader and the primary author. Cecilia
Tam and SimonBennett, successive co-ordinators
of the Energy Technology Roadmaps programme
and Jean-FranoisGagn, Head of the Energy
Technology Policy Division, were responsible for the
development of this roadmap and provided valuable
input and leadership throughout. PhilippeBenoit
and DidierHoussin provided important guidance.
PierpaoloCazzola, AraceliFernandesPales and
UweRemme provided significant input and support.
Many other IEA colleagues provided thoughtful
comments and support including FranoisCuenot,
TimurGl, YoshikiEndo, SteveHeinen,
SimonMller, LuisMunuera, CdricPhilibert,
DanielePoponi, YasuhiroSakuma and TaliTrigg.
This work was developed in collaboration with
governments, industry, experts, and the IEA energy
technology network. The roadmap was supported
by the Japanese Ministry of Economy, Trade, and
Industry (METI), NOW GmbH-National Organisation
Hydrogen and Fuel Cell Technology, Adam Opel
AG, Air Liquide, BP plc, Honda R&D Co Ltd, Nissan
Motor Co, Shell, Toyota Motor Co, US Department
of Energy (US DOE), the European Commission, the
Advanced Fuel Cell Implementing Agreement and
the Hydrogen Implementing Agreement.
The volunteers of the Hydrogen and Fuel Cell
Technology Roadmap steering committee
have provided guidance over the course of its
development: RosemaryAlbinson, Michael Ball,
HannoButsch, AndrzejChmura, AndyFuchs,
NancyGarland, RittmarvonHelmolt,
HitoshiIgarashi, HiroyukiKanesaka, EricMiller,
PhilippeMulard, TadashiNomura, EijiOhira,
HideoOkamoto and MarcSteen.
Finally, the IEA would like to thank the numerous

experts who provided the authors with information
and/or comments on drafts of the roadmap
and Technology Annex: FlorianAusfelder,
LucasBertrand, RobertBienenfeld, JeanPierreBirat, TudorConstantinescu, BenoitDecourt,
EricDenhoff, PeterDevlin, JavierDufour,
RayEaton, AmgadElgowainy, KariEspegren,
LewFulton, RobertFriedland, MontereyGardiner,
MichaelGasik, JohnGerman, JongheeHan,
RobHarvey, TakuyaHasegawa, ArneHll,
RobertJudd, HiroyukiKaneko, TimKarlsson,
AndreasKopp, JunboomKim, J.Leaver,
ShannaKinghts, JochenLinssen, PaulLucchese,
MorryMarkowitz, AngeloMoreno, HarukiMotegi,
NishimuraMotohiko, YujiNagata, JoanOgden,
YoshimiOkada, ToddOnderdonk, BengtRidell,
ValentinoRomeri, JulienRoussel, JacquesSaintJust, SejiSano, MasaharuSasakura, RichardSmithBingham, YuichiSone, DetlefStolten, PaulTanaka,
AndreiTchouvelev, ChihiroTobe, MaryRosedeValladares, ElliVarkaraki, ManfredWaidhas,
ToshioWaku, KazunoriWatanabe, JrgWind,
MarcelWeeda and TiejunZhao.
The authors would also like to thank JustinFrenchBrooks for editing the manuscript a well as the
IEA publication unit, in particular Astrid Dumond,
ThereseWalsh, Rebecca Gaghen and Bertrand Sadin
for their assistance in layout and graphical design
support.
For more information on this document, contact:
Technology Roadmaps
9, rue de la Fdration
75739 Paris Cedex 15
France
Email: TechnologyRoadmapsContact@iea.org
Acknowledgements
Key findings
Cross-cutting opportunities
offered by hydrogen and
fuel cells
zz H
ydrogen is a flexible energy carrier that can be
produced from any regionally prevalent primary
energy source. Moreover, it can be effectively
transformed into any form of energy for diverse
end-use applications. Hydrogen is particularly
well suited for use in fuel cells that efficiently use
hydrogen to generate electricity.
zz H
ydrogen with a low-carbon footprint has the
potential to facilitate significant reductions in
energy-related CO2 emissions and to contribute
to limiting global temperature rise to 2C,
as outlined in the high hydrogen variant
(2DShighH2) of theIEAEnergy Technology
Perspectives (ETP) 2C Scenario (2DS). In addition,
hydrogen use can lower local air pollutants and
noise emissions compared to direct fossil fuel
combustion. By enabling continued use of fossil
fuel resources for end-use applications under a
2DS, hydrogen production in combination with
CCS can provide energy security benefits and
help maintain a diversified fuel mix.
zz A
s an energy carrier, hydrogen can enable new
linkages between energy supply and demand,
in both a centralised or decentralised manner,
potentially enhancing overall energy system
flexibility. By connecting different energy
transmission and distribution (T&D) networks,
sources of low-carbon energy can be connected
to end-use applications that are challenging
to decarbonise, including transport, industry
and buildings. In remote areas with little access
to the power grid, these connections can
expand off-grid access to energy services while
minimising emissions.
of hydrogen can be stored over long periods of

time, facilitating the integration of high shares of
variable renewable energy (VRE) into the energy
system for power and heat. Hydrogen-based
systems such as power-to-fuel, power-to-power
or power-to-gas can be employed to make use of
VRE that would otherwise be curtailed at times
when supply outstrips demand.
zz F CEVs can provide the mobility service of todays
conventional cars at potentially very low-carbon
emissions. Deploying a 25% share of FCEVs in
road transport by 2050 can contribute up to
10% of all cumulative transport-related carbon
emission reductions necessary to move from an
ETP6C Scenario (6DS) to a 2DS, depending on
the region. Assuming a fast ramp-up of FCEV
sales, a self-sustaining market could be achieved
within 15 to 20 years after the introduction of the
first 10000 FCEVs.
zz W
hile the potential environmental and energy
security benefits of hydrogen and fuel cells
in end-use applications are promising, the
development of hydrogen generation, T&D and
retail infrastructure is challenging. For example,
the risks associated with market uptake of FCEVs
have been a significant barrier to infrastructure
investment. For each of the assumed
150millionFCEVs sold between now and 2050,
around USD900 to USD1900 will need to be
spent on hydrogen infrastructure development,
depending on the region.1
1. Unless otherwise stated, all monetary values are in 2013 USD.
Energy storage and

utilisation in transport,
industry and buildings
zz H
ydrogen is particularly useful as an energy
carrier, because it allows low-carbon energy
to be stored. Small quantities of hydrogen
with low-carbon footprint can be stored under
restricted space and weight requirements to
enable long-distance, low-carbon driving using
fuel cell electric vehicles (FCEVs). Large quantities

Cross-cutting opportunities
offered by hydrogen and
fuel cells
Energy storage and

utilisation in transport,
industry and buildings
zz E ncourage fuel efficiency and low greenhouse gas

emission technologies across all energy sectors
through market driven, technology- and fuelneutral policies. A stable policy and regulatory
framework including for example carbon
pricing, feed in tariffs, fuel economy standards,
renewable fuel standards or zero-emission vehicle
mandates is important for raising market
certainty for investors and entrepreneurs.
zz P
rove on-road practicality and economics across
the supply chain of FCEVs by putting the first
tens of thousands of vehicles on the road, along
with hydrogen generation, T&D and refuelling
infrastructure, including at least 500 to 1000
stations in suitable regions around the world, and
cross-border projects. Build upon deployment
programmes in Europe, Japan, Korea and
California as well as the use of captive fleets.
zz S
timulate investment and early market
deployment of hydrogen and fuel cell
technologies and their infrastructure through
effective policy support to bring down costs.
National and regional priorities should determine
the value chains and the market barriers to be
targeted.
zz E ngage international stakeholders from relevant

industries as well as regional, national and
local authorities in developing risk-mitigation
strategies, including the development of financial
instruments and innovative business models that
de-risk hydrogen T&D and retail infrastructure
development for FCEV market introduction.
zz C
ontinue to strengthen and harmonise
international codes and standards necessary
for safe and reliable handling and metering of
hydrogen in end-use applications.
zz Increase the number of hydrogen-based energy

storage systems suitable for integrating VRE and
collect and analyse performance data under reallife conditions.
zz K
eep up supporting technology progress and
innovation by unlocking public and private funds
for RD&D for key hydrogen technologies, such as
fuel cells and electrolysers. Enhance the focus on
cross-cutting research areas, such as materials,
that could play a transformative role in improving
performance. Where possible, promote projects
with international cooperation to maximise the
efficiency of funding.
zz E stablish regulatory frameworks that remove

barriers to grid access for electricity storage
systems including power-to-fuel and powerto-gas applications. Where regionally relevant,
establish a regulatory framework for the blending
of hydrogen into the natural gas grid.
zz Improve understanding of regionally specific

interactions between different energy sectors
through integrated modelling approaches to
quantify benefits of energy system integration.
zz W
here regionally relevant, accelerate activities
directed at developing the capture and storage of
CO2 from fossil-derived hydrogen production into
mature business activities.
zz Increase data on resource availability and costs

for hydrogen generation at national and regional
levels. Analyse the potential future availability of
curtailed electricity for hydrogen production as a
function of VRE integration, other power system
flexibility options and competing demands for
any surplus renewable electricity.
zz A
ddress potential market barriers where
opportunities exist for the use of low-carbon
hydrogen in industry (e.g. in refineries).
zz E xtend information campaigns and educational
programs to increase awareness-raising.
Introduction
Hydrogen is a flexible energy carrier with potential
applications across all energy sectors. It is one of
only a few potential near-zero emission energy
carriers, alongside electricity and advanced
biofuels. Nonetheless, it is important to note
that hydrogen is an energy carrier and not an
energy source: although hydrogen as a molecular
component is abundant in nature, energy needs to
be used to generate pure hydrogen. The hydrogen
can then be used as a fuel for end-use conversion
processes, for example using fuel cells to produce
power. As is the case for electricity generation,
hydrogen production incurs a cost and suffers from
thermodynamic losses.
Hydrogen can be produced from various primary or
secondary energy sources, depending on regional
availability. Primary energy sources useful for
hydrogen production comprise renewable sources,
such as biomass, and also fossil fuels, such as
natural gas and coal. Electricity can also be used for
hydrogen generation using electrolysers, which are
a pivotal technology for enabling the splitting of
water into its components hydrogen and oxygen.
Hydrogen itself contains no carbon if used in a
fuel cell or burned in a heat engine, water or water
vapour is the only exhaust. Nevertheless, hydrogen
can have a very significant carbon footprint. Its
lifecycle carbon emissions are determined by the
primary energy source and the process used for
hydrogen production, and need to be taken into
account when quantifying climate benefits.
While not ignoring the implications of hydrogen
generation pathways, this roadmap focuses
primarily on the demand side of the energy system.
There, hydrogen could play an important role
in future road transport, as FCEVs can be a lowcarbon alternative to conventional passenger cars
and trucks. In buildings, micro co-generation
units could increase energy efficiency. 2 In the
longer run, industrial processes in the refining,
steel and chemical industries could be substantially
decarbonised through the use of hydrogen with a
low-carbon footprint. In many, but not all of these
applications, fuel cells are an important technology
for converting hydrogen to power and heat. Fuel
cells are intimately but not exclusively linked to
hydrogen. They can also be used with other fuels
such as natural gas or even liquid hydrocarbons, thus
helping their early adoption.
2. Co-generation refers to the combined production of heat

and power.
Producing hydrogen from electricity and storing

it in gaseous or liquefied form could be an option
for increasing the flexibility of the energy system,
allowing for the integration of high shares of VRE.
Hydrogen can enable power-to-x trajectories its
capability of being converted to various forms of final
energy, such as power, heat and transport fuels, can
be used to join subsystems of the energy system that
historically had no, limited or only one-way linkages.
This is what makes this roadmap especially
challenging. Many of the technology components
are less mature than technologies featured in
other IEA Technology Roadmaps, adding greater
uncertainty to technological and economic
parameters. A proper inter-sectoral view of the
energy system also requires integrated modelling,
which becomes highly complex if the target is
significant temporal and spatial detail. For this
roadmap the IEA ETP toolbox has been enhanced to
account for some of the synergies that emerge when
high shares of VRE integration on the energy supply
side are combined with demand for hydrogen as
a fuel.
Rationale for hydrogen

and fuel cell technologies
As outlined in the 2015 edition of ETP (IEA, 2015),
contributions to reducing GHG emissions from the
energy supply sector and all energy demand sectors
will be needed if dangerous climate change is to be
prevented.
On the energy supply side, the power sector needs
to be deeply decarbonised if an ambitious emission
reduction scenario to limit global warming to 2C
above pre-industrial levels is to be achieved. On a
global scale, annual emissions need to be reduced
by 85% by 2050 compared to todays levels, which is
achieved in the 2DS to a large extent by an increase
of renewable power to about 63% of generated
electricity by 2050. This high level of renewable
energy integration, which following the 2DS will
need to be exceeded in certain regions such as the
EuropeanUnion, necessitates a deep structural
change in the way we operate power systems.
Discussion of low-carbon energy systems frequently
centres on issues such as flexibility and system
integration. Todays perception of flexibility is
mostly related to energy supply. In fact, it is closely
linked to energy storage. Fossil resources store
immense amounts of energy. They can be used
when and where necessary, their high energy

density (either in gaseous, liquid or solid form)
allowing them to be efficiently transported over
long distances. This inherently provides the energy
system with a lot of flexibility. In a low-carbon
energy system based on high shares of VRE, this
temporal and spatial flexibility to modulate energy
supply according to demand is limited.
Electricity from VRE carries the temporal and spatial
imprint of its resource: sunlight, wind, tidal and
wave patterns. Their patterns are not necessarily
aligned with variations in demand with regard
both to location and time of supply. This causes
periods of supply surplus and deficit, which
will differ from one place to another. Moreover,
fluctuating output as a result of weather variability
can lead to rapid swings in supply. This is a
challenge, because the electricity grid requires
electricity supply and demand to be in balance
instantaneously and at all times. A suite of options is
available to overcome the space and time mismatch
of variable electricity supply and demand. Grid
infrastructure, flexible generation, demand-side
response and energy storage can all be used in this
way, but should be used according to their relative
economic performance.
Hydrogen generated from electricity and water can
be stored in large quantities over long periods and
re-transformed to electricity (power-to-power)
although at an efficiency cost of more than 70% of
the input electricity. It can be mixed into the natural
gas grid or converted to synthetic methane powerto-gas) or sold as fuel for FCEVs to the transport
sector (power-to-fuel). Hydrogen may thus open up
entirely new ways to integrate renewable electricity
in the energy system and compensate in part for
the loss of flexibility resulting from reduced use of
fossil fuels.
Decoupling energy use and carbon emissions on
the energy supply side needs to be complemented
by measures within energy demand sectors,
notably transport, buildings and industry. The main
mitigation options are technological improvement
(either through efficiency improvements of
conventional technologies or the deployment of
new technologies) and behavioural change to
reduce energy use, as well as switching to lowcarbon fuels.
Road transport is a large carbon emitter, accounting

for about three-quarters of all transport emissions.
Apart from avoiding road transport demand and
shifting it to more efficient transport modes,
such as passenger and freight rail, substantially
decarbonising the road transport sector can be
achieved by: 1) increasing the share of direct use of
low-carbon electricity via battery electric vehicles
(BEVs) and plug-in hybrid electric vehicles (PHEVs);
2) significantly raising the share of sustainable
low-carbon biofuels in combination with highefficiency hybridised internal combustion engine
(ICE) vehicles and PHEVs; 3) the use of FCEVs fuelled
by low-carbon footprint hydrogen. All three options
can substantially contribute to reducing emissions,
but hinge on overcoming different barriers. Energy
storage is again pivotal the higher the demand
for autonomy, the greater the need for energy to be
stored on board.
BEVs can draw upon existing electricity generation
and T&D infrastructure, and rely on the fact that
their carbon impact would be reduced by the
decarbonisation already taking place in the power
sector. Still, batteries face a serious trade-off
between energy capacity and weight, and range
anxiety and recharging time are major concerns
for consumers. In the case of biofuels, production
raises doubts with respect to sustainability and
displacement of food production, particularly as
considerable amounts of biofuels will be necessary
to decarbonise long-haul road freight, aviation
and shipping. By contrast, FCEVs could provide
transport utility comparable to todays vehicles
while, at the same time, meeting climate and energy
security targets. Here, the challenge is to build up
an entirely new hydrogen generation, T&D and
retail network. The main barrier to overcome is the
risk related to committing investment in large-scale
FCEV production on the one hand, and hydrogen
infrastructure roll-out on the other, particularly
against a background of high uncertainty with
respect to FCEV market uptake. Therefore, a better
understanding of consumer preferences with
regard to vehicle range, refuelling and recharging
infrastructure provision as well as safety concerns is
key to improve projections of the market potential
of low or zero-emission vehicles.
Introduction
Figure1: Energy system today and in the future

Future
Today
H2
Heat network
Electricity grid
Liquid and gaseous fuels and feed-stocks T&D
Hydrogen
KEY POINT: Hydrogen can link different energy sectors and energy T&D networks and thus increase
the operational flexibility of future low-carbon energy systems.
Substituting fossil-derived hydrogen with

low-carbon footprint hydrogen in industrial
applications also offers significant potential
for carbon emission mitigation. Globally, the
refining, chemical and industrial gas industries
use approximately 7.2exajoules (EJ) of hydrogen
per year (Suresh etal., 2013). Around 48% of this
is currently produced from natural gas, using
steam methane reforming (SMR) without CCS,
30% arises as a fraction of petroleum during the
refining process, 18% is produced from coal, and
the balance (4%) is electrolytic hydrogen (Decourt
et al., 2014). Altogether, the used hydrogen
resulted in annual emissions of approximately
500megatonnes(Mt)CO2. In general, depending
on region-specific natural gas prices, hydrogen
produced via large-scale SMR processes is typically
available at relatively low costs. This together
with anticipated T&D costs will set the benchmark
against which alternative, low-carbon hydrogen
production pathways need to be measured.
In the steel industry, more efficiently integrating the
hydrogen generated in classic blast furnaces in the
steelmaking process could deliver significant carbon
emission reductions today. Processes to directly
reduce iron ore (DRI) in the presence of hydrogen
10
could unlock an important mitigation potential,

especially if low-carbon footprint hydrogen was
available at competitive cost.
A schematic representation of todays energy
system and a potential low-carbon energy system of
the future are shown in Figure1. The key difference
lies in the different energy vectors used to supply
transport, buildings and industry, and in particular
the T&D of electricity, heat, and liquid as well as
gaseous fuels via different energy networks. Todays
energy system is heavily dependent on fossil fuels
and, apart from co-generation, few connections
exist between the different T&D systems. In a
future system, hydrogen could play a pivotal role
by connecting different layers of infrastructure in a
low-carbon energy system.
The use of hydrogen as an energy carrier is
closely linked to the deployment of fuel cells and
electrolysers. Fuel cells are the key technology
to efficiently convert hydrogen into electricity
to propel FCEVs, or for using it in other end-use
applications in buildings or industry (eventually
exploiting the waste heat for heating purposes). In
addition, fuel cells can also convert a range of other
hydrocarbon fuels, such as natural gas or methanol,
and the immediate use of such fuels for which there
is existing infrastructure in fuel cells could be an

important step to help reduce technology costs that
remain high today.
Electrolyser technology is pivotal to establish the
new links between the power sector and transport,
buildings and industry. They allow the conversion of
renewable electricity into hydrogen, a zero carbon
chemical fuel and feedstock, by splitting water into
hydrogen and oxygen.
Purpose, process and

structure of the roadmap
The purpose of this roadmap is to lay out
hydrogens potential in different energy sectors,
and also its limitations. The roadmap aims to:
zz P
rovide an extensive discussion of the nature,
function and cost of key hydrogen technologies.
zz Identify applications where using hydrogen can

offer the maximum added value.
zz Identify the most important actions required in
the short and long term to successfully develop
and deploy hydrogen technologies in support of
global energy and climate goals.
zz Increase understanding among a range of
stakeholders of the potential offered by hydrogen
technologies, particularly the synergies they offer
existing energy systems.
This roadmap was developed with the support
of a wide range of stakeholders, including
members of industry, academia and government
institutions. To facilitate collaboration, the IEA
Hydrogen Technology Roadmap team hosted three
regional expert workshops to examine regionspecific opportunities for and barriers to hydrogen
technology deployment (Table1).
Table1: W
orkshops parallel to the development
of the Technology Roadmap on Hydrogen and Fuel Cells
Date
Workshop focus
9-10July 2013
Kick-off meeting and Europe-focused expert workshop: scope, technology, market

and policy discussion
28-29January 2014
NorthAmerica-focused expert workshop: hydrogen generation pathways, technology,

market and policy discussion
26-27June 2014
Asia-focused expert workshop: technology, market and policy discussion
Due to the roadmaps broad scope, covering both

energy supply and several energy demand sectors,
the detailed results provided in the Vision section
focus on selected regions, including EU4 (France,
Germany, Italy and the UnitedKingdom), Japan and
the UnitedStates.
Roadmap scope
zz h
ydrogen in the energy supply sector VRE
integration and energy storage, comprising
power-to-power, power-to-gas and power-to-fuel
zz h
ydrogen infrastructure T&D, storage and retail
technologies
zz k ey hydrogen generation and conversion
technologies electrolysers and fuel cells.
The following applications are the focus of this

roadmap:
zz h
ydrogen-based systems in energy demand
sectors FCEVs in transport, fuel cell micro cogeneration in the residential sector and selected
applications in the refining, steel and chemical
industries
Introduction
11

In 2013, global hydrogen usage amounted to a total
of 7.2EJ (Suresh et al., 2013). To date this hydrogen
has not been used as an energy carrier, i.e. it is
not converted into electricity, mechanical energy
or heat to be used for energy service. Hydrogen is
almost entirely used as feedstock within the refining
and chemical industries to convert raw materials
into chemical or refinery products.
The generation of hydrogen from fossil resources, its
transmission, distribution and use within industry
and the refining sector are based on mature
technologies and applied on a large scale, and are
not the main focus of this roadmap. However, these
mature technologies will play an important role in a
transition to low-carbon hydrogen.
The use of hydrogen as an energy carrier is
beginning to emerge although the first FCEVs were
developed in the 1960s, it is only in the last ten years
that the technology has developed to an extent
that certain car manufacturers are announcing the
launch of FCEVs. Toyota launched its Mirai (Future)
model in Japan in 2014, Hyundai is planning to
begin the sale of FCEVs in the near future (the
Hyundai Tucson FCEV has been available for lease
since summer 2014), and Honda announced plans to
launch its next generation FCEV in 2016. Although
predicted production numbers are a small fraction
of conventional passenger car sales, or even those of
electric vehicles, they show the increased interest of
car manufactures in this technology.3
A similar trend can be observed in the field of
energy storage applications. Increasing numbers
of hydrogen-based large-scale energy storage
demonstration projects are being launched,
planned or announced, with a remarkable
concentration of activity in Germany, motivated by
the attempt to explore benefits for the integration
of VREs. Likewise, opportunities to store large
amounts of hydrogen using chemical hydrides are
being actively explored in Japan.
Japan certainly leads the field in the stationary

application of fuel cell technology, with more than
120000Ene-farm domestic fuel cell micro cogeneration systems already installed (NEDO, 2014).
In the following sections, technology status and
opportunities are reviewed for hydrogen and fuel
cell applications and individual technologies.
FCEVs together with hydrogen T&D and retail
infrastructure, hydrogen-based energy storage
systems, hydrogen technologies in industry and
fuel cells in buildings are considered in turn. This
is followed by a more detailed discussion of some
of the key technologies for generating, using and
storing hydrogen.
An overview of hydrogen systems in the transport
sector and their techno-economic parameters is
shown in Table 2. More detailed technical data on
hydrogen technology components, such as fuel
cells and electrolysers, are briefly discussed in the
sections Key hydrogen production technologies
and Key hydrogen conversion technologies as
well as in the Roadmap Technology Annex.
Although other pathways to use hydrogen as a fuel
in transport are feasible, e.g. via the use of synthetic
methane in compressed natural gas (CNG) vehicles
or through conversion to methanol, the current
analysis focuses on FCEVs and the use of pure
hydrogen.
3. In 2013 around 63millionpassenger light-duty vehicles

(PLDVs) were sold globally (OICA, 2014), while in 2014 around
300000BEVs and PHEVs were sold (EVI, 2015).
12
Table 2: Current performance of hydrogen systems in the transport sector

Application
Power or energy
Energy efficiency*
capacity
Investment
cost**
Lifetime
Maturity
Fuel cell vehicles
80 - 120kW
Tank-to-wheel
efficiency
43-60% (HHV)
USD60000100000
150000km
Early market
introduction
Hydrogen retail
stations
200kg/day
~80%, incl.
compression to
70MPa
USD1.5million2.5million
Early market
introduction
Tube trailer
(gaseous) for
hydrogen delivery
Up to 1 000kg
~100% (without
compression)
USD1 000 000

(USD 1 000 per
kg payload)
Mature
Liquid tankers for

hydrogen delivery
Up to 4000kg
Boil-off stream:
0.3% loss per day
USD750000
Mature
* Unless otherwise stated, efficiencies are based on lower heating values (LHV).
** All power-specific investment costs refer to the energy output.
Notes: HHV = higher heating value; kg = kilogram; kW = kilowatt.
Sources: IEA data; Decourt et al. (2014), Hydrogen-Based Energy Conversion, More than Storage: System Flexibility; Elgowainy (2014),
Hydrogen infrastructure analysis in early markets of FCEVs, IEA Hydrogen Roadmap North America Workshop; ETSAP (2014),
Hydrogen Production and Distribution; Iiyama et al. (2014), FCEV Development at Nissan, ECS Transactions, Vol. 3, pp. 11-17; Nexant
(2007), Liquefaction and pipeline costs, Hydrogen Delivery Analysis Meeting, 8-9May; NREL (2014), Hydrogen Station Compression,
Storage and Dispensing - Technical Status and Costs; NREL (2012a), National Fuel Cell Electric Vehicle Learning Demonstration Final
Report; USDOE (2010a), Hydrogen Program 2010 Annual Progress Report - Innovative Hydrogen Liquefaction Cycle; USDOE (2010b), DOE
Hydrogen Program 2010 Annual Progress Report - Technology Validation Sub-Program Overview; Yang and Ogden (2007), Determining
the lowest-cost hydrogen delivery mode, International Journal of Hydrogen Energy, pp. 268-286.
FCEVs
FCEVs are essentially electric vehicles using hydrogen
stored in a pressurised tank and a fuel cell for onboard power generation. FCEVs are also hybrid
cars, as braking energy is recuperated and stored
in a battery. The electric power from the battery
is used to reduce peak demand from the fuel cell
during acceleration and to optimize its operational
efficiency. Being both electric and hybrid vehicles,
FCEVs benefit from technological advancement
in both technologies, since they have a significant

amount of parts such as batteries and power
electronics in common (McKinsey and Co., 2011).
Today around 550FCEVs (passenger cars and buses)
are running in several demonstration projects
across the world (Table 3). A small number of fuel
cell heavy freight trucks (HFTs) are currently being
used in a demonstration project at the port of
LosAngeles, testing the usability of range extenders
with electric trucks.
Table 3: Existing FCEV fleet and targets announced by hydrogen initiatives

Country or region
Running FCEVs
Planned FCEVs on the road

2015
2020
Europe
192
5000
~350000
Japan
102
1000
100000
Korea
100
5000
50000
UnitedStates
146
~300
~20000
Sources: Weeda et al. (2014), Towards a Comprehensive Hydrogen Infrastructure for Fuel Cell Electric Cars in View of EU GHG Reduction
Targets; personal contact with US Department of Energy; Japanese registration number from database of Japan Automobile Dealers
Association (JADA, March, 2015).
13
To date, FCEVs are fuelled with gaseous hydrogen

at pressures of 35MPa to 70Mpa. As 70MPa tanks
allow for much higher ranges at acceptable tank
volumes, most recent demonstration vehicles are
equipped with these.
conventional cars and refuelling time is about the

same, FCEVs can provide the mobility service of
conventional cars at much lower carbon emissions,
depending on the hydrogen generation pathway
(Figure 2).
Currently, on-road fuel economy is around 1kg of

hydrogen per 100km travelled, and demonstration
cars have ranges of around 500km to 650km. Since
the driving performance of FCEVs is comparable to
Vehicle costs remain high FCEV prices announced

to date have been set at around USD60000
(Toyota, 2015) during the early market introduction
phase. Announced prices might rather reflect the
Figure 2: W
ell-to-wheel (WTW) emissions vs. vehicle range
for several technology options
180
ICE
2013
WTW CO2 emissions (gCO2/km)
160
FCEV
140
2013
BEV
120
2013
BEV
100
2013
80
2050
60
40
2050
20
PHEV
2050
2050
0
0
200
400
600
800
1 000
1 200
1 400
Range (km)
Notes: gCO2/km = grams carbon dioxide per kilometre; WTW = wheel-to-wheel; the upper range of BEV emissions takes into account
todays average world power generation mix, the lower range is based on 100% renewable electricity; the upper range of FCEV
emissions takes into account a. hydrogen production mix of 90% NG SMR and 10% grid electricity, the lower range is based on 100%
renewable hydrogen; the lower range of PHEV emissions takes into account 65% electric driving; by 2050, a biofuel share of 30% is
assumed for PHEVs and ICEs.
KEY POINT: FCEVs can achieve a mobility service compared to todays conventional
cars at potentially very low WTW carbon emissions.
assumed customers willingness to pay than the

costs to produce the vehicles. Current FCEV models
are targeted at high-income and technophile
early movers living close to hydrogen refuelling
infrastructure clusters, which are starting to develop
in California, Germany, Japan and Korea.
The high cost of the fuel cell systems is driving
total vehicle costs, and the current challenge lies in
reducing fuel cell stack and balance of plant (BOP)
costs while simultaneously increasing lifetime.
While economies of scale have huge potential to
14
drive down fuel cell costs, the cost of the highpressure tank is largely determined by expensive
composite materials, which are expected to fall
much more slowly (Argonne National Laboratory
- Nuclear Division, 2010). This is why the focus
of recent R&D has been on accelerating cost
reductions in composite materials for high-pressure
tanks. To bring down the costs of the entire
FCEV, manufacturers are currently focusing on
technology packaging, to finally be able to mount
the fuel cell power train on the same chassis used
for conventional cars.
To realise their full performance potential against

conventional cars, FCEVs target the medium
and upper size car segments. Initially, costly
technologies are typically introduced in premium
cars, but in the longer term more than threequarters (vehicle class C and higher [IEA (2012)])
of the passenger light-duty vehicle (PLDV) market
would be suitable for fuel cell technology.
Since FCEVs will target the same vehicle class like
plug-in hybrids medium and upper size class
vehicles able to cover large distances these might
be the closest competing low-carbon technology.
Compared to plug-in hybrids, FCEVs could enable
very low-emission individual motorised transport.
At high annual production rates and under
optimistic assumptions with regard to fuel cell
systems and hydrogen storage tanks, FCEVs have
the potential to be less costly than plug-in hybrids.
This is largely due to their lower complexity since
they do not require two different drive-trains.
Fleet vehicles can play a significant role in the initial
market introduction phase. Refuelling at a base
location allows the necessary hydrogen refuelling
infrastructure, and the associated costs, to be kept
to a minimum. As a result of better utilisation of the
refuelling equipment and higher annual mileages,
economic viability of fleet FCEVs could be achieved
earlier than for individually owned vehicles. The
French HyWAY programme, for example, aims to derisk the development of infrastructure for FCEVs by
focusing on captive fleets.
Broad personal vehicle ownership of FCEVs may also
hinge upon overcoming consumer concerns about
passenger safety in collisions, ability of the general
public to safely refuel, and safety in tunnels or
enclosed parking spaces.
Heavy-duty vehicles such as trucks and buses
can also be equipped with fuel-cell powertrains.
Significant experience with fuel cell buses already
exists (McKinsey and Co., 2012) and partly results
from being able to draw upon the fleet vehicle
advantage. Public transport subsidies are common
and could ease the introduction of fuel cell
technology in that field. Furthermore, co-benefits
such as reduced air pollution can be an important
argument for FCEV and particularly fuel-cell bus
deployment, especially in heavily polluted and
densely populated urban areas around the world.
Fuel cell trucks are one of only very limited
options available to deeply decarbonise heavyduty, long-haul road freight transport. Although
competition with other low-carbon technologies
is less pronounced in that segment, fuel cell

long-haul HFTs will face difficulty competing with
advanced conventional trucks. HFT diesel engines
can already achieve high efficiencies (up to 40%)
during constant highway cruising speeds. Fuel
cell efficiencies decline with increasing power
output, and using them in HFTs decreases the
efficiency benefit compared to conventional
technology. Furthermore, as HFTs require longrange autonomy, on-board storage of the necessary
volumes of hydrogen becomes critical. Compared to
conventional diesel technology, hydrogen stored at
70MPa still needs four times more space to achieve
the same range, even taking into account the higher
efficiency of the fuel cell powertrain (IEA, 2012).
The potential role of fuel cell technology in HFTs is
thus more uncertain.
Hydrogen T&D
Hydrogen refuelling stations can be supplied by
one of two alternative technologies: hydrogen
can be produced at the refuelling station using
smaller-scale electrolysers or natural gas steam
methane reformers, or can be transported from a
centralised production plant. Each approach has its
own advantages and trade-offs. While large-scale,
centralised hydrogen production offers economies
of scale to minimise the cost of hydrogen generation,
the need to distribute the hydrogen results in higher
T&D costs. Meanwhile, the opposite is true for
decentralised hydrogen generation. While T&D costs
are minimised, smaller-scale production adds costs at
the hydrogen generation stage. Finding the optimal
network configuration requires detailed analysis
taking into account the full range of local factors,
such as geographic distribution of resources for
hydrogen production, existing hydrogen generation
and T&D infrastructure, anticipated hydrogen
demand at the retail station and distance between
the place of hydrogen production and hydrogen
demand. However, economies of scale realised in
large centralised hydrogen generation facilities tend
to potentially outweigh the additional costs of longer
T&D distances.
A number of options are available for hydrogen
T&D: gaseous truck transport; liquefied truck
transport; and pumping gaseous hydrogen through
pipelines (Table 4). A trade-off exists between fixed
and variable costs: while gaseous truck delivery
has the lowest investment cost, variable costs are
high as a result of the lower transport capacity. The
opposite is true for pipelines fixed costs are driven
by high investment costs. Once the pipeline is fully
15
Table4: Q
ualitative overview of hydrogen T&D technologies for hydrogen
delivery in the transport sector
Capacity
Transport
distance
Energy loss
Fixed costs
Variable
costs
Deployment
phase
Gaseous tube
trailers
Low
Low
Low
Low
High
Near term
Liquefied truck
trailers
Medium
High
High
Medium
Medium
Medium to
long term
High
High
Low
High
Low
Medium to
long term
Hydrogen
pipelines
utilised, the variable costs are low. The lowest-cost

pathway depends on many factors, with hydrogen
demand at the refuelling station and T&D distance
being the most important.
Hydrogen refuelling stations

Hydrogen refuelling stations are a critical element
in the fuel supply chain for FCEVs, as providing
a minimum network density is a prerequisite to
attaining consumer interest. They can be exclusively
for hydrogen or part of a multi-fuel station.
The set-up of a hydrogen station is largely
determined by daily hydrogen demand, the form
of hydrogen storage on board the vehicle (e.g. the
pressure and the phase), and the way hydrogen is
delivered to or produced at the station. Determining
the optimal size of a station is a critical step. While

very small stations with daily capacities of 50kg
to 100kg of hydrogen might be necessary in the
beginning (basically allowing for 10 to 20refills a
day), stations up to 2000kg per day will be needed
in a mature market.
The link to hydrogen T&D technologies is obvious.
While small stations could be based on gaseous
trucking or on-site hydrogen production, liquefied
trucking or the use of pipelines are the only options
for hydrogen delivery to stations larger than 500kg
per day, if the hydrogen is not produced on-site.
The set-up of the station hence implies a certain
path dependency, which complicates investment
decision-making, as multiple risks (mainly linked
to the pace of FCEV market uptake and hydrogen
demand) need to be taken into account (Box1).
Box1: Risks associated with investment in hydrogen refuelling stations

The investment risk associated with the
development of refuelling stations is mainly
due to high capital and operational costs, and
the under-utilisation of the facilities during
FCEV market development, which can lead
to a negative cumulative cash flow over 10 to
15years (Figure 3).
This long valley of death can be minimised
by reducing capital and operation costs and
maximising asset utilisation. High capital costs
are mainly linked to hydrogen compression and
storage. The higher the pressure of hydrogen
16
stored on board FCEVs, the more expensive

are the compressors needed at the station
a 35MPa refuelling station is about one
third less costly than a 70MPa station. The
requirement for compression at the station can
be minimised either by delivering the hydrogen
at high pressure from the hydrogen generation
plant, or by lowering the required pressure on
board the FCEV. It would even be possible to
provide hydrogen at several pressure levels,
where only stations on long-distance corridors
would provide the option to refuel at 70MPa.
Cumulative annual cash flow

of a hydrogen refueling station network
Figure 3: C
umulative cash flow curve of hydrogen stations
in the early market phase
Measures for optimising the business case:
Reduction of investment costs
Reduction of operational expenses
Improvement of utilisation
Public support
Valley of Death
10 to 15 years
KEY POINT: The valley of death can last for 10 to 15 years

until positive cumulative cash flow is achieved
Clustering hydrogen stations around main
demand centres and connecting corridors
during the FCEV roll-out phase can ensure
maximising utilisation rates. Several
partnerships and initiatives, such as the
California Fuel Cell Partnership (CaFCP) in
the United States, H2Mobility in Europe or
the Fuel Cell Commercialisation Conference
of Japan (FCCJ), have made proposals for the
optimal roll-out of hydrogen stations so as to
provide maximum coverage at minimal cost.
An overview of existing and planned hydrogen
refuelling stations is given in Table 5.
To cover the negative cash flow period, direct

public support might be needed for hydrogen
stations during the FCEV market introduction
phase. As recently published in a paper by
Ogden et al, direct subsidies in the range of
USD 400 000 to USD 600 000 per station
might be required until cumulative cash flow
becomes positive, assuming a fast uptake of the
Californian FCEV market (Ogden, J. et al, 2014).
Table 5: E
xisting public hydrogen refuelling stations
and targets announced by hydrogen initiatives
Country or region
Existing hydrogen refuelling stations
Planned stations
2015
2020
Europe
36
~80
~430
Japan
21
100
>100
Korea
13
43
200
UnitedStates
>50
>100
Sources: Weeda et al. (2014), Towards a Comprehensive Hydrogen Infrastructure for Fuel Cell Electric Cars in View of EU GHG Reduction Targets;
HySUT (2014), Fuel Cell Vehicle Demonstration and Hydrogen Infrastructure Project in Japan; FFC (2015), Fuel Cell Commercialisation Conference
in Japan (FCCJ), http://fccj.jp/hystation/index.html#hystop; personal contact with US Department of Energy.
17
Box 2: Carbon footprint of hydrogen used in transport

The carbon footprint for different hydrogen
pathways and for gasoline and diesel are shown
in Figure 4 for the European Union. Depending
on the production and T&D pathway,
todays carbon footprint for hydrogen can be
significant. Decentralised hydrogen production
(at the refuelling station) using todays EU grid
electricity mix, and including compression to
88 MPa, results in a carbon footprint which is
almost three times higher than that for gasoline
or natural gas. Conversely, when produced
from renewable power, biomass or fossil fuels
with CCS, the carbon content of hydrogen can
be reduced to below 20 gCO2eq per MJ. Still,
in combination with the higher efficiency of
FCEVs, the use of hydrogen from natural gas
SMR without CCS results in lower per kilometre
emissions than the use of gasoline in comparably
sized conventional cars (see also Figure 2).
to the energy-intense compression of the

hydrogen gas to 88 MPa, but also due to
hydrogen T&D using trucks (with hydrogen
either in gaseous or liquefied form) or
pipelines. The values shown in Figure 4 are
for the European Union and contain relatively
long transmission distances for natural gas
of 4 000 km (Transportation to market).
Since transmission distances might be shorter
in the United States, carbon footprint values
could be slightly lower, while LNG supply in
Japan would lead to higher specific carbon
emissions. Furthermore, the comparison
suggests that the liquefaction of hydrogen for
T&D purposes leads to around 25% to 30%
higher carbon emission compared to gaseous
truck or pipeline transport.
In the future, the carbon footprint of lowcarbon hydrogen could be reduced further
if low-carbon electricity was used for
compression.
Hydrogen T&D and retailing (Conditioning

and distribution) have a substantial carbon
emission contribution, which is mainly due
Figure 4: T
odays carbon footprint for various hydrogen pathways and for
gasoline and compressed natural gas in the European Union
Decentralised electrolysis, grid electricity, compression 880 bar
Centralised electrolysis, wind electricity, pipeline T&D, compression 880 bar
Centralised NG SMR, pipeline T&D, compression 880 bar
Centralised NG SMR, gaseous truck T&D, compression 880 bar
Centralised NG SMR, liquefaction, liquid truck T&D, compression 880 bar

Centralised NG SMR with CCS, pipeline T&D, compression 880 bar
50
100
150
200
250
gCO2eq per MJ hydrogen

Production & conditioning at source
Transportation to market
Transformation near market
Conditioning & distribution
Source: adapted from JRC (2013), Technical Reports Well-to-tank Report Version 4.0 JEC Well-to-Wheels Analysis, Joint
Research Centre, Publication Office of the European Union, Luxembourg.
KEY POINT: Depending on the generation, T&D and retail pathway, the carbon footprint of
hydrogen can vary between almost 20 and more than 230gCO2per MJ.
18
Hydrogen for VRE integration

The integration of large shares of VRE into the
energy system will go hand-in-hand with the need
to increase the operational flexibility of the power
system. This implies the need to store electricity
that is not needed at the time or the place of
generation, or to transform it in a way that it can be
used in another sector of the energy system.
A wide range of options and strategies exist to
integrate high shares of variable generation (in
the order of 30% to 45% in annual electricity
generation) cost-effectively without the use of

large-scale seasonal storage (IEA, 2014c). However,
taking into account the full range of local conditions
(such as regulatory and market structure, the
status of existing and planned grid infrastructure
investments) when analysing potential deployment
opportunities for energy storage technologies,
or attaining even higher VRE shares while also
achieving the 2DS, can imply a greater need to
apply such storage technologies at large scale.
Table 6: C
urrent performance of hydrogen systems
for large-scale energy storage
Application
Power or
energy
capacity
Energy
efficiency*
Investment cost**
Lifetime
Maturity
Power-to-power
(including
underground
storage)
GWh to
TWh
29% (HHV, with

alkaline EL) 33% (HHV, with
PEM EL)
1900 (with alkaline EL) 6300 USD/kW (with PEM

EL) plus ~8 USD/kWh for
storage
20000 to
60000hours
(stack lifetime
electrolyser)
Demonstration
Underground
storage
GWh to
TWh
90-95%, incl.
com-pression
~8 USD/kWh
30years
Demonstration
Power-to-gas
(hydrogenenriched natural
gas, HENG)
GWh to
TWh
~73% excl. gas

turbine (HHV)

EL), excl. gas turbine
20000 to
60000hours
(stack lifetime
electrolyser)
Demonstration
20000 to
60000 hours
(stack lifetime
electrolyser)
Demonstration
Power-to-gas
(methanation)
~26% incl. gas

turbine (PtP)
GWh to
TWh
~58% excl. gas

turbine (HHV)
~21% incl. gas
turbine (PtP)

EL), incl. gas turbine (PtP)
EL), excl. gas turbine
EL), incl. gas turbine (PtP)
* = Unless otherwise stated, efficiencies are based on LHV.

** = All investment costs refer to the energy output.
Notes: excl. = excluding; incl. = including; PtP = power-to-power; GWh = gigawatt hour; TWh = terawatt hour.
Source: IEA data; Decourt et al. (2014), Hydrogen-based energy conversion. More than Storage: System Flexibility; Giner Inc. (2013), PEM
electrolyser incorporating an advanced low-cost membrane, 2013 Hydrogen Program Annual Merit Review Meeting; ETSAP (2014),
Hydrogen Production and Distribution; Hydrogen Implementing Agreement Task 25 (2009), Alkaline Electrolysis; NREL (2009b), Scenario
Development and Analysis of Hydrogen as a Large-Scale Energy Storage Medium; Saur (2008), Wind-To-Hydrogen Project: Electrolyzer
Capital Cost Study; Schaber, Steinke and Hamacher (2013), Managing temporary oversupply from renewables efficiently: Electricity
storage versus energy sector coupling in Germany, International Energy Workshop, Paris; Stolzenburg et al. (2014), Integration von
Wind-Wasserstoff-Systemen in das Energiesystem Abschlussbericht; USDOE (2010b), Hydrogen Program 2010 Annual Progress Report
Technology Validation Sub-Program Overview; USDOE (2014b), Hydrogen and Fuel Cells Program Record.
19
Electricity storage systems can be classified by size

according to their input and output power capacity
(megawatts [MW]) and their discharge duration
(hours). These three parameters finally determine
energy capacity (MWh). Together with the expected
annual number of cycles, round-trip efficiency and
self-discharge, the annual full-load hours can be
determined. Location within the energy system and

response time are other important parameters (see
also the Technology Roadmap on Energy Storage
[IEA, 2014b]). Hydrogen-based technologies
are best suited to large-scale electricity storage
applications at the megawatt scale, covering hourly
to seasonal storage times (Figure 5).
Figure 5: Electricity storage applications and technologies
10 MW
1 MW
100 kW
CAES
Flywheel
H2
T&D
100 MW
Offgrid
utility
scale
End-user
10 kW
Offgrid/
end-user
self cons.
Day
Microsecond Second Minute Hour
Week
Discharge duration
PHS
Generation
1 GW
Battery
Arbitrage
Seasonal
Inter-seasonal storage
storage
Supercapacitors
Small-scale
wind PV:
grid support
T&D deferral
Frequency
regulation
Black start
Load following
Voltage
regulation
Power requirment
Large-scale
wind PV:
grid support
Siting
Technology
Applications
1 kW
Season
Day
Microsecond Second Minute Hour
Week
Discharge duration
Season
Note: CAES = compressed air energy storage; PHS = pumped hydro energy storage.
KEY POINT: Hydrogen-based electricity storage covers large-scale and long-term storage applications.
However, hydrogen-based systems to integrate

otherwise-curtailed electricity are not restricted
to electricity storage only. As mentioned before,
hydrogen-based energy storage systems could
be used to integrate surplus VRE electricity across
different energy sectors, e.g. as a fuel in transport or
as a feedstock in industry. They can be categorised
as follows:
zz P
ower-to-power: electricity is transformed
into hydrogen via electrolysis, stored in an
underground cavern or a pressurised tank and
re-electrified when needed using a fuel cell or a
hydrogen gas turbine.
zz P
ower-to-gas: electricity is transformed into
hydrogen via electrolysis. It is then blended in the
natural gas grid (hydrogen-enriched natural gas
20
HENG) or transformed to synthetic methane in a

subsequent methanation step. For methanation, a
low-cost CO2 source is necessary.
zz P
ower-to-fuel: electricity is transformed into
hydrogen and then used as a fuel for FCEVs in the
transport sector.
zz P
ower-to-feedstock: electricity is transformed into
hydrogen and then used as a feedstock, e.g. in
the refining industry.
All hydrogen-based VRE integration pathways
are based on several transformation steps, which
finally lead to rather low efficiencies over the whole
conversion chain in the range of 20% to 30%
(Figure6). It is important to only compare final
energies of the same quality, for example electricity
either used in the power system or on board

FCEVs. The greater the number of conversion steps
included, the lower the overall efficiency.
The trade-off between power-to-power and
power-to-gas options lies within the higher overall
efficiency of pure power-to-power applications
versus the possibility of using existing storage and
T&D infrastructure for power-to-gas systems. The
latter might be a strong argument in the near term
otherwise-curtailed renewable electricity could

be integrated into the energy system via blending
hydrogen to up to 5% to 10% in the natural gas
mix, or transforming it directly to synthetic natural
gas via methanation. Although no compatibility
issues with subsequent end-use technologies arise
in case of power-to-gas including methanation, the
poor overall efficiency is likely to pose a substantial
barrier for deployment.
Figure 6: C
urrent conversion efficiencies of various hydrogen-based
VRE integration pathways
Power-to-power
100
Electricity
73
Electrolysis
67
Compression
Power-to-gas (blending)
Electricity
100
Power-to-power
Electrolysis
Power-to-gas (methanation)
ElectroElectricity
100
lysis
73
Compression
70
58
Compression
73
Methanation
73
Compression
T&D
55
68
T&D
Gas turbine
54
26
Power-to-power
Gas
21
turbine
Power-to-power
Power-to-fuel
Electricity
29
Fuel cell
100
Electrolysis
67
T&D
64
Retail
54
Fuel cell
24
Note: The numbers denote useful energy; except for gas turbines, efficiencies are based on HHV; the conversion efficiency of gas
turbines is based on LHV.
KEY POINT: Total round-trip efficiencies of hydrogen-based energy storage applications are low.
In an electricity system with high levels of VRE, it

can be expected that supply will outstrip demand
in some periods of the day and year. This has been
labelled excess or surplus electricity. While
some situations might be envisaged in which
the storage operator incurs no costs, curtailed
VRE electricity is generated at the same costs as
VRE electricity required by the system, i.e. the
consumer will pay for curtailed electricity through
higher per unit electricity prices, since capital
costs need to be recovered by selling less output
electricity than in the case with no curtailment.
Nevertheless, in this case, conversion efficiency has
no impact on levelised costs of the final energy.
However, using otherwise-curtailed VRE power to
generate hydrogen poses an economic challenge
for multiple reasons. Firstly, electrolysers have
significant investment costs, which means that
they will only be cost effective if they are operated
for a sufficient amount of time during the year. As
periods of surplus VRE generation will occur only

for a limited amount of time, relying exclusively on
generation surpluses is likely to be insufficient to
reach sufficient capacity factors. Hence, it is likely
that electricity with at least some value will be used
for hydrogen production. Secondly, each conversion
step on the way from electricity to hydrogen and
back to electricity entails losses (Figure 6). Losses
are of minor importance if the input electricity
cannot be used for other applications, i.e. it would
otherwise need to be curtailed. However, hydrogen
generation will compete with other possible uses
of surplus electricity, such as thermal storage.
These challenges point to two areas for technology
improvement: increasing efficiencies and reducing
investment costs.
Only focusing on improving the technology is not
sufficient; new and more integrated approaches
need to be applied to create viable business cases.
21
As for all long-term, large-scale energy storage

systems, annual full-load hours are limited. While
technology components such as electrolysers and
fuel cells remain expensive, all possible energy
system services or by-products need to be exploited
to the fullest extent possible, adopting the benefits
stacking principle (IEA, 2014b).
When using electrolysers and fuel cells, a number
of by-products, such as oxygen (during electrolysis)
or process heat are produced, which need to be
sold separately or used on site. In case of powerto-power systems, it is beneficial not only to
sell power generated from low-value, surplus
electricity, but also to provide ancillary services and
to take part in the power control market. Here, the
provision of controllable negative and positive load
is remunerated.
Participating in different energy markets can
help to create profits. Bi-generation (hydrogen
and electricity) or even tri-generation systems
(hydrogen, electricity and heat) offer the possibility
of selling their products at the respective highest
price, i.e. electricity and heat during times of peak
demand and hydrogen to the transport sector,
depending on the market conditions.
Large-scale underground
hydrogen storage
Storing hydrogen-rich gaseous energy carriers
underground has a long history and became
popular with the use of town gas to provide energy
for heating and lighting purposes in the middle of
the nineteenth century.
A geological formation can be suitable for hydrogen
storage if tightness is assured, the pollution of
the hydrogen gas through bacteria or organic
and non-organic compounds is minimal, and the
development of the storage and the borehole is
possible at acceptable costs. Actual availability
of suitable geological formations where energy
storage is required is another limiting factor.
Comparing different underground storage
options with respect to safety, technical feasibility,
investment cost and operational cost, using salt
caverns currently appears to be the most favourable
option (Table 7), being already deployed at several
sites in the UnitedStates and the UnitedKingdom.
Table 7: Q
ualitative overview of characteristics of geological formations
suitable for hydrogen storage
Salt caverns
Depleted oil
fields
Depleted
gas fields
Aquifers
Lined rock Unlined rock

caverns
caverns
Safety
++
Technical feasibility
++
++
++
Investment costs
++
Operation costs
++
++
Source: adapted from HyUnder (2013), Assessment of the Potential, the Actors and Relevant Business Cases for Large Scale and Seasonal
Storage of Renewable Electricity by Hydrogen Underground Storage in Europe - Benchmarking of Selected Storage Options.
Box 3: Power-to-gas in Europe: storage potential and limitations

Most developed countries have extensive
natural gas T&D networks, including significant
natural gas underground storage in depleted
oil and gas fields, and salt caverns. This existing
infrastructure offers huge energy storage
22
potential if hydrogen produced from otherwisecurtailed renewable electricity was blended into
natural gas (HENG). For example, the EU natural
gas grid accounts for more than 2.2 million km
of pipelines and about 100 000 million cubic
metres of natural gas (which equals roughly

1100 TWh) can be stored in dedicated storage
sites (Eurogas, 2014). Assuming a volumetric
blend share of 5% hydrogen in the natural gas,
a theoretical storage potential of around 15
TWh of hydrogen (or roughly 9 TWh of output
electricity) could be available using the existing
natural gas storage infrastructure. If all natural
gas used during a year in Europe was blended at
the same share, more than 60 TWh of hydrogen
(roughly equalling 36 TWh of output electricity)
could be integrated in the energy system.
Second, the much lower volumetric energy

density of hydrogen compared to natural gas
significantly reduces both the energy capacity
and efficiency of the natural gas T&D system
at higher blend shares. At 20% volumetric
blend share, flow rate needs to be increased
by around 10% to provide the same energy to
the customer, and pipeline storage capacity
needed to balance intra-day fluctuations
decreases by 20% (Decourt et al., 2014). By far
the strongest restriction is set by compression
stations and various end-use applications
connected to the gas grid. According to a
recent German study (Deutscher Verein des
Gas- und Wasserfaches, 2013), compressing
stations, gas turbines and CNG tanks (e.g. in
CNG vehicles) currently restrict acceptable
blend shares to 2% by volume without any
further adjustment (Figure 7).
Blending hydrogen into the natural gas grid

faces several limitations. First, the ability of
hydrogen to embrittle steel materials used for
pipelines and pipeline armatures necessitates
upper blending limits of around 20% to
30%, depending on the pipeline pressure
and regional specification of steel quality.
Figure 7: Limitations on the blend share of hydrogen by application

70%
60%
50%
40%
30%
20%
10%
0%
Further research
needed
Transport
Distribution
Measuring and control
Gas stoves
Co-generation plants
Fuel cells
Stirling motor
Fan burners
Condensing boilers
Gas burners
Engines
CNG tanks
Odorisation
Pressure regulation
Gas chromatographs
Quantity transformers
Turbine meters
Diaphragm meters
Ultrasonic meters
Valves
House installation
Gas flow detectors
Seals
Connections
Steel pipelines
Plastic pipelines
Tanks
Storage
Storage installations
Pore storages
Cavern storages
Gas turbines
Compression stations
Transmission pipelines
Adjustment &
modification
needed
H2 blending
uncritical
Appliances
Source: Deutscher Verein des Gas- und Wasserfaches (2013), Entwicklung von Modularen Konzepten zur Erzeugung, Speicherung
und Einspeisung von Wasserstoff und Methan in Erdgasnetz.
KEY POINT: The most critical applications with respect to the blend share
of hydrogen are gas turbines, compressing stations and CNG tanks.
A recent article in Hydrogen Energy (Gahleitner,

2013) provides a good overview of the technical
and economical parameters of almost 50 powerto-gas pilot plants (of which the majority remain
in operation), concluding that apart from the
technical core components, design and size as
well as control strategy and system integration
have a significant influence on overall system

efficiency. Unsurprisingly, efficiency, cost,
reliability and lifetime of electrolysers are the
main areas where improvement is needed. To
date, few of the pilot plants have been operated
for lengthy periods.
23
Box 3: Power-to-gas in Europe: storage potential and limitations (continued)

In the near term, the potential of power-to-gas
applications to contribute to VRE integration
might be constrained to specific locations
fulfilling a suite of prerequisites. It requires
the local availability of significant amounts of
otherwise-curtailed renewable power and an
existing natural gas infrastructure with wellknown end-use applications. Blend shares
of up to 10% of hydrogen might be viable
in local natural gas distribution networks,
if modifications to gas turbines located
downstream were applied and no CNG cars
were supplied.
Hydrogen in industry
Most of todays hydrogen demand is generated and
used on industrial sites as captive hydrogen. In the
EU more than 60% of hydrogen is captive, one-third
is supplied from by-product sources, and less than
10% of the market is met by merchant hydrogen
(Kopp, A., 2013). In general, industrial hydrogen
demand offers a significant potential for carbon
emission mitigation, but the cost of low-carbon
hydrogen is critical.
Hydrogen in the refining industry

Most of the hydrogen used in the refining industry
is used for hydro-treating, hydro-cracking and
desulphurisation during the refining process. A
steadily growing demand for high-quality, lowsulphur fuels, together with a decline in light and
sweet crude oils, is leading to a growing demand
for hydrogen. In the past, most of the required
hydrogen was produced on site from naphtha,
which itself is a refinery product, using catalytic
reformation (Rabiei, 2012). Matching the hydrogen
balance is becoming increasingly difficult, and
therefore the oil refining industry is using more
hydrogen from natural gas steam reformation, most
often produced in large dedicated plants managed
by industrial gas companies.
Under business-as-usual assumptions, it is estimated
that by 2030 more than twice the amount of
hydrogen will be used in the refining sector
compared to 2005 (IFP Energies Nouvelles, 2008).
Most of the growth is expected to take place in
24
The summary report of the recent workshop

titled Putting Science into Standards:
Power-to-Hydrogen and HCNG held at the
Joint Research Centre (JRC) of the European
Commission in Petten, concludes, amongst
others, that setting a clear limit for blending
hydrogen into natural gas is currently seen
as premature. It underlines the need for
harmonisation of current and future standards
with regard to the allowed hydrogen content
in gas mixtures, and points out at CNG vehicle
tanks to be a main bottleneck for HENG
application (JRC, 2014).
North and SouthAmerica, where the impact of

using super-heavy crudes and crude oil from oil
sands is most significant. China could see a tripling
in hydrogen demand in the refinery sector.
In addition to conventional fuel refining, the
upgrading of second-generation, sustainable
biofuels produced from lignocellulosic biomass
might also demand considerable amounts of
hydrogen for hydro-deoxygenation in the future.
The decarbonisation of hydrogen can therefore
have a significant impact on reducing the carbon
footprint of conventional fuels and biofuels during
the refining process.
Considerable experience in transmitting hydrogen
via pipeline already exists. In the UnitedStates the
existing hydrogen pipeline system amounts to
some 2400km, while in Europe almost 1600km
are already in place (Pacific Northwest National
Laboratory, 2015).
Hydrogen in the steel industry

Hydrogen is generated in the steel industry as
part of by-product gases during the coke, iron and
steelmaking processes. For the most part, these
off-gases are used to contribute to on site thermal
requirements.
Currently, 71% of steel production is based on the
reduction of iron ore in conventional blast furnaces
(World Steel Association, 2014) where coke, coal
and/or natural gas are used as reducing agents.
The resulting pig iron is then reacted with oxygen
in a basic oxygen furnace in order to remove excess

carbon content from the iron and to generate
liquid steel.
reducing agent within the process. Compared to a

typical blast furnace, coke demand per tonne of pig
iron has been significantly reduced.
Hydrogen-containing gases are generated during

coke production (coke oven gas, COG), and also
in the blast furnace (blast furnace gas, BFG) and
the basic oxygen furnace (basic oxygen furnace
gas, BOFG).4 These gas streams globally represent
around 8.0EJperyear and can displace other fossil
fuels for heating purposes once collected and treated
for reuse on site. In 2012, around 68% were reused
in iron and steel production processes; alternatively
these gases are flared.
In Japan, the process developed under the COURSE

50 research project (CO2 Ultimate Reduction in
Steelmaking Process by Innovative Technology
for Cool Earth 50) enables the introduction of
hydrogen-enriched COG into the blast furnace to
reduce carbon emissions. This research project also
aims to separate and recover CO2 from the BFG. The
Korean consortium POSCO/RIST is also developing
a conversion process to produce a hydrogen-rich
gas from COG and CO2 through steam reforming,
which could be used for iron ore reduction in a blast
furnace or SR process.
The more efficient use of by-product hydrogen

during the steelmaking process can contribute
to improved overall energy efficiency and hence
reduced carbon emissions.
In order to minimise the need for investment in
dedicated hydrogen production plants, by-product
hydrogen could also be used as a fuel for FCEVs
during the early stages of market introduction.
However, purification and cleaning of the hydrogen
gas necessary for further use in proton exchange
membrane fuel cells (PEMFCs) is economically
challenging.
Hydrogen-rich gases can also be used as a reducing
agent in alternative methods of steel production.
Both, the DRI process and the smelt reduction (SR)
process allow the production of iron without the
need for coke. As coke production is very carbon
intensive, important emission reductions can be
achieved when the whole process chain is assessed.
In DRI processes, further emission reductions are
feasible with the use of hydrogen with a low-carbon
footprint. Instead of using natural gas as reducing
agent, hydrogen produced from fossil fuels with CCS
or renewable electricity could significantly reduce
carbon emissions, if available at competitive costs.
Several research programmes, such as the
European-based Ultra-Low-Carbon Dioxide
Steelmaking (ULCOS), have focused on improving
the performance of DRI and SR processes and
exploring alternatives to optimise the use of
process gas streams as iron ore reducing agents.
Within these programmes, alternative blast furnace
arrangements have been developed that collect,
treat and reuse the blast furnace top gas as a
4. These gases typically contain hydrogen in the range of 39% to
65% by volume for COG, 1% to 5% by volume for BFG and 2% to
10% by volume for BOFG (EuropeanCommission, 2000).
Fuel cell technology

in buildings
The co-generation of power and heat allows the
waste heat that occurs during power generation to
be used for heating purposes. This can significantly
increase overall energy efficiency in the buildings
sector. Decentralised generation of electricity and
heat using micro co-generation systems enables
this benefit to be realised in the absence of district
heating networks. Many different natural gaspowered co-generation systems using ICEs are
already available on the market.
Fuel cell micro co-generation systems powered by
natural gas are an alternative to conventional ICE
systems. Currently, the electrical efficiency of fuel
cell micro co-generation systems is around 42%,
being around 10percentage points higher than for
ICE micro co-generation systems. The downside is
significantly higher investment cost: while ICEbased systems cost around USD2200perkW,
commercially available fuel cell systems typically
cost more than USD9000perkW for commercial
applications (Pacific Northwest National Laboratory,
2013) and more than USD 18000perkW for home
systems (Hydrogen and Fuel Cell Strategy Council,
2014; IEA AFC IA, 2014).
Fuel cell micro co-generation systems are either
based on a PEMFC or a solid oxide fuel cell (SOFC),
the latter providing much higher temperature heat.
Although systems with up to 50kW electrical output
exist, most commercially available systems have
electrical power outputs of around 1kW, therefore
being insufficient to fully supply the average US
or European dwelling. However, in the Japanese
25
Table 8: Current performance of fuel cell systems in the buildings sector

Application
Fuel cell micro
co-generation
Power or
energy capacity
Energy
efficiency*
Investment
cost**
0.3-25kW
Electric:
35-50% (HHV)
<20000 USD/kW
(home system, 1 kWe)
Co-generation:
up to 95%
<10000USD/kW
(commercial system, 25 kWe)
Life
time
Maturity
60000- Early market

90000 introduction
hours
* = Unless otherwise stated efficiencies are based on LHV.

Notes: 1 kWe = kilowatt electric output.
Source: Pacific Northwest National Laboratory (2013), Business Case for a Micro Combined Heat and Power Fuel Cell System in Commercial
Applications; Hydrogen and Fuel Cell Strategy Council (2014), Strategic Roadmap for Hydrogen Fuel Cells; IEA AFC IA (2014), IEA AFC IA
Annex Meeting 25.
market more than 120000Ene-Farm fuel cell micro

co-generation systems of that power category have
already been sold under a government subsidy that
lasted until September 2014.
All natural gas-based micro co-generation systems
need a high difference between local natural gas
and electricity prices, the so-called spark-spread.
Together with higher efficiency, annual availability

and government incentives, the spark-spread
forms the economic basis for selecting a micro
co-generation system over grid electricity and
conventional domestic hot water boilers for heating
and hot water supply.
Box 4: The Japanese Ene-Farm experience

In 2009, a consortium of major Japanese
energy suppliers and fuel cell manufacturers
began marketing co-branded fuel cell micro
co-generation units with an electrical output
of between 700W and 1000W to Japanese
customers. The Ene-Farm system can be ordered
with two different fuel cell types, using PEMFC
and SOFC technologies, with PEMFC systems
making up 90% of cumulative sales. With power
output up to 1kW, the system is not intended
to cover the entire electricity demand of an
apartment or family house, but to significantly
contribute to the electricity demand and to fully
cover the hot water demand.
Since their introduction, approximately
120000 units have been installed in Japanese
buildings (Figure 8). Initially a subsidy of
almost USD15000 per unit was granted by
the Japanese government, dropping to below
USD4000 by 2014. Overall, the unit price
had fallen from around USD45000 in 2009 to
around USD19000 by 2014. This means that a
learning rate of more than 15% (i.e. reduction in
26
price per doubling of installed units) has been

achieved during the large-scale demonstration
phase.* The Ene-Farm system demonstrates that
similar learning rates assumed for PEMFC units
in the car-manufacturing sector are feasible at
larger annual production volumes.
Nonetheless, to reach the target of several
hundred thousand installed units in the near
future, and millions of units by 2030 (as suggested
by the Ene-Farm consortium), costs need to
decline even further. So far, around 30% of the
cost of the PEMFC unit is accounted for by the
water tank (which is necessary for conventional
boilers as well) and only 15% by the fuel-cell
stack. The BOP (25%), the fuel-processing unit
(15%) and the packaging (15%) account for the
remaining 55% of total costs (Hydrogen and Fuel
Cell Strategy Council, 2014). This means that
further cost reductions will be harder to achieve,
as the fuel stack is currently a relatively small share
of the overall cost of the unit.
* To calculate the learning rate, year 2010 with almost 10000
units installed has been taken as the baseline.
Figure 8: E
ne-Farm fuel cell micro co-generation cumulative sales,
subsidies and estimated prices, 200914
140 000
50 000
45 000
120 000
35 000
30 000
80 000
USD
100 000
Units
Cumulative units
installed
40 000
Estimated price
without subsidy
25 000
60 000
20 000
15 000
40 000
10 000
20 000
Subsidy
5 000
0
2009
2010
2011
2012
2013
2014
Sources: Hydrogen and Fuel Cell Strategy Council (2014), Strategic Roadmap for Hydrogen Fuel Cells; IEA AFC IA (2014), IEA AFC IA
Annex Meeting 25.
KEY POINT: The price of Ene-Farm fuel cell micro co-generation systems
has fallen by more than 50% since 2009.
In Japan, equipping 10% of households (i.e. 5.3

million) with fuel-cell co-generation systems is
estimated to cut total residential energy demand
by 3%, resulting in 4% emission reductions
compared to the use of gas boilers and grid
electricity for residential energy supply.
Other niche applications

based on fuel cell
technologies
Several other hydrogen-based niche applications
exist that are currently applied across different
sectors. These applications comprise fuel cell
powered fork lifts, autonomous power systems for
either stationary or portable off-grid applications,
and uninterruptible power systems for back-up
power.
Since 2009, over 8200 fuel cell materials handling
equipment units have been deployed in the
UnitedStates. Benefitting from longer lifetimes and
shorter and less-frequent refuelling cycles, fuel cell
forklifts have demonstrated acceptable payback
Ene-Farm products are also intended to be sold

on the European market, where differences
in gas quality and much higher presence of
potentially poisonous constituents in the
European gas require the re-engineering of the
gas processing unit for PEM systems.
periods and improved cost-effectiveness compared

to battery-powered forklift applications used in
indoor warehouse operations (US DOE, 2014c).
Stationary fuel cell systems in the range of several
kilowatts to multiple megawatts are used for remote
power and back-up power applications. They are
used to supply for example telecommunication
towers, networking equipment or datacentres with
resilient and reliable power. In these cases, fuel cell
systems often replace diesel generators, providing
longer lifetimes as well as less maintenance. The
entire range of fuel cell types is represented within
this market. While smaller systems in the range up to
several kilowatts of output electricity are most often
based on PEMFCs, bigger systems up to the multimegawatt range mostly build on high-temperature
fuel cells such as molten carbonate (MC) or solid
27
oxide (SO) fuel cells. Many of the fuel cell systems

rely on natural gas or hydrogen as primary fuel,
but other liquid fuels such as methanol, ethanol,
liquefied petroleum gas (LPG) and diesel or kerosene
as well as gaseous fuels such as biogas, propane,
butane and coal syngas are being used as well. In
2013, stationary fuel cell systems accounted for
almost 90% of the shipped systems (US DOE, 2014a).
Key hydrogen
generation technologies
The following sections briefly discuss selected
hydrogen generation technologies, such as
reformers and electrolysers. The Technical Annex to
this roadmap provides more detailed information
on specific technical issues.
Steam methane reforming

Around 48% of hydrogen is currently produced
from natural gas using the SMR process, which is
based on a reaction of methane and water steam at
high temperatures in the presence of a catalyst. As
CO2 concentration in the exhaust gas is high, SMR
units are promising candidates for the application
of CCS technology, potentially leading to an 80%
reduction in its carbon emissions.
Produced on a large scale, hydrogen costs mainly
depend on the natural gas price, and are currently
between USD0.9perkg in the UnitedStates,
USD2.2perkg in Europe and USD3.2perkg in
Japan. 5 Very small-scale reforming units exist with
production rates down to 4.5kg of hydrogen per
hour, but generation costs are much higher and in
the same order of magnitude as hydrogen produced
via electrolysis (Table9).
5. Based on IEA calculations taking into account current natural gas
prices of USD13perMWh in the UnitedStates, USD37perMWh
in the EuropeanUnion and USD56perMWh in Japan.
Table 9: Current performance of key hydrogen generation technologies

Power or
capacity
Ffficiency*
Initial
investment cost
Life time
Maturity
Steam methane
reformer, large
scale
150-300 MW
70-85%
400-600 USD/kW
30 years
Mature
Steam methane
reformer, small
scale
0.15-15 MW
~51%
3 000-5 000 USD/kW
15 years
Demonstration
Alkaline
electrolyser
Up to 150 MW
65-82% (HHV)
850-1 500 USD/kW
60 00090 000 hours
Mature
PEM electrolyser
Up to 150 kW
(stacks)
Up to 1 MW
(systems)
65-78% (HHV)
1 500-3 800 USD/kW
Lab scale
85-90% (HHV)
Application
SO electrolyser
20 000Early market
60 000 hours
~1 000 h
R&D

Notes: PEM = proton exchange membrane; SO = solid oxide.
Sources: IEA data; Decourt et al. (2014), Hydrogen-Based Energy Conversion, More Than Storage: System Flexibility; ETSAP (2014), Hydrogen
Production and Distribution; FCH-JU (2014), Development of Water Electrolysis in the European Union, Fuel Cells and Hydrogen Joint
Undertaking; Giner Inc. (2013), PEM electrolyser incorporating an advanced low-cost membrane, 2013 Hydrogen Program Annual Merit
Review Meeting; Hydrogen Implementing Agreement Task 25 (2009), Alkaline Electrolysis; IKA RWTH Aachen (n.d.), On-site Hydrogen
Generators from Hydrocarbons, www.ika.rwth-aachen.de/r2h/index.php/On-site_Hydrogen_Generators_from_Hydrocarbons; Linde
(n.d.), Hydrogen, www.linde-engineering.com/internet.global.linde engineering.global/en/images/H2_1_1_e_12_150dpi19_4258.pdf;
NREL (2009a), Scenario development and analysis of hydrogen as a large-scale energy storage medium, RMEL Meeting; Saur (2008),
Wind-To-Hydrogen Project: Electrolyzer Capital Cost Study; Schaber, Steinke and Hamacher (2013), Managing temporary oversupply from
renewables efficiently: electricity storage versus energy sector coupling in Germany, International Energy Workshop, Paris; Stolzenburg
et al. (2014), Integration von Wind-Wasserstoff-Systemen in das Energiesystem Abschlussbericht; US DOE (2014b), US DOE (2010a),
Hydrogen Program 2010 Annual Progress Report - Innovative Hydrogen Liquefaction Cycle.
28
28
Reforming processes are not limited to the use of

natural gas. All hydrogen-rich gases can be used
to produce pure hydrogen via adapted reforming
processes. Following gasification as a first step,
hydrogen can be produced from other fossil
resources such as coal and also from biomass or
organic waste materials.
For electrolysers using only electric power (and

no external heat) as input energy, the efficiency of
hydrogen production decreases with cell voltage
while the hydrogen production rate increases with
cell voltage. At a given cell geometry, the operator
therefore has to deal with a trade-off between
electrolyser efficiency and hydrogen output.
Electrolysis
Different types of electrolysers are distinguished

by their electrolyte and the charge carrier, and can
be grouped into: 1) alkaline electrolysers; 2) PEM
electrolysers; and 3) SO electrolysers.
Electrolysis is a process of splitting water into

hydrogen and oxygen by applying a direct current,
converting electricity into chemical energy.
Currently, around 8GW of electrolysis capacity are
installed worldwide (Decourt et al., 2014).
Figure 9: S
chematic representation of technology development potential
of different electrolysers
Alkaline (commercial)
2.0
PEM (commercial) and

advanced alkaline (R&D)
Cell voltage (V)
Higher efficiency
1.5
Solid oxide (R&D)

1.0
0.5
Usual operating range of current density
0
0.5
1.5
1.0
2.0
Electrolysis current density (A/cm )

Lower capital cost
Note: A/cm = ampere per square centimetre.
Source: adapted from Decourt et al. (2014), Hydrogen-Based Energy Conversion, More Than Storage: System Flexibility.
KEY POINT: Although alkaline electrolysers are a mature and affordable technology, PEM and SO
electrolysers show a greater potential to reduce capital costs and to increase efficiency.
All electrolysers consist out of the electrolyser stack,

comprising up to 100cells, and the BOP. Stacks
can be mounted in parallel using the same BOP
infrastructure, which is why electrolysers are highly
modular systems. While this makes the technology
very flexible with respect to hydrogen production
capacity, it also limits the effects of economies
of scale, as even large electrolysers are based on
identically sized cells and stacks.
29
Alkaline electrolysers are currently the most

mature technology, and investment costs are
significantly lower than for other electrolyser types.
Although alkaline electrolysers currently have
higher efficiencies than electrolysers using solid
electrolytes, PEM and SO electrolysers have much
higher potential for future cost reduction and, in
case of SO electrolysers, efficiency improvements
(Figure 9). PEM electrolysers are particularly
29
interesting as they show both the highest current

density and operational range, prerequisites
necessary to reduce investment costs and improve
operational flexibility at the same time. As of today,
cell lifetime is a limiting factor for PEM and SO
electrolyser technologies.
The cost of electrolytic hydrogen is largely
determined by the cost of electricity and the
investment costs associated with the electrolyser.
Minimising the costs of input electricity is likely to
be accompanied by lower annual utilisation rates,
as very low-cost, surplus renewable electricity
will only be available for a limited amount of time
per year, which further stresses the impact of
investment costs. It is therefore important to find
the right balance between reducing investment
costs and achieving efficiency improvements.
Key hydrogen conversion

and storage technologies
The following section briefly discusses selected
hydrogen conversion and storage technologies.
The Technical Annex to this roadmap provides more
detailed information on specific technical issues.
Fuel cells
Fuel cells allow the oxidation of hydrogen-rich
fuel and its conversion to useful energy without
burning it in an open flame. Compared to other
single-stage processes to convert chemical energy
into electricity, e.g. open-cycle gas turbines, their
electrical efficiency is higher and in the range of
32% to up to 70% (HHV).
Figure 10: Production volumes of fuel cells according to application

200
40
150
30
100
20
Capacity additions
50
10
Transportation
Thousand units
MW
Stationary
Portable
Units by application
Stationary
Transportation
2008
2009
2010
2011
2012
2013
Portable
Source: USDOE (2014a), 2013 Fuel Cell Technologies Market Report.
KEY POINT: Currently, more than 80% of all fuel cells sold are used in stationary applications.
Fuel cells operate with a variety of input fuels,

not only hydrogen. These include natural gas and
also liquid fuels such as methanol or diesel. If pure
hydrogen is used, the exhaust of fuel cells is water
vapour and so has very low local environmental
impact. However, if hydrocarbon fuels are used,
using fuel cells for power generation produces CO2
emissions, and so can only confer a climate benefit
by operating at higher efficiency than alternative
combustion methods. Nevertheless, experience
with fuel cells based on hydrocarbons has a
high value for low-carbon innovation due to the
applicability of technological advances to fuel cells
more generally. This is partly because hydrocarbon
fuels are often reformed to hydrogen in a step
30
that precedes the fuel cell and also because some

hydrocarbons may be produced by lower carbon
processes in future, e.g. methanol.
Similar to electrolysers, fuel cells are subject to a
trade-off between efficiency and power output.
Efficiency is highest at low loads and decreases
with increasing power output. In comparison to
conventional technologies, fuel cells can achieve
their highest efficiencies under transient cycles,
such as in passenger cars.
As in the case of electrolysers, different fuel cell
types exist, which can mainly be distinguished by
their membrane type and operating temperature.
Fuel cells can be categorised into: 1) PEMFC;

2) alkaline fuel cell; 3) phosphoric acid fuel cell
(PAFC); 4) molten carbonate fuel cell (MCFC); and
5) SOFC. While PEMFCs and alkaline fuel cells have
low operating temperatures of around 80C, the
others operate at higher temperatures of up to
600C (SOFC), which makes them more suitable

to combined heat and power applications. The
higher the temperature, the better the efficiency at
otherwise similar parameters. PEMFCs are the most
suitable option for FCEVs.
Figure 11: P
roduction cost for PEMFCs for FCEVs
as a function of annual production
USD/kW
350
US DOE
300
2007 estimate
250
2014 estimate
200
2020 target
150
Ultimate target
100
Power trains for Europe
50
High
0
0
100 000
200 000
300 000
Annual production
400 000
500 000
Low
Sources: adapted from McKinsey and Co. (2011), A Portfolio of Powertrains for Europe: a Fact-Based Analysis, The Role of Battery Electric
Vehicles, Plug-in Hybrids and Fuel Cell Electric Vehicles; US DOE (2012), Fuel Cell Technologies Program Record; US DOE (2014d), DOE Fuel
Cell Technologies Office Record Fuel Cell System Costs.
KEY POINT: Although current PEMFC systems for FCEVs cost around USD300 to USD500 per kW,
cost can be reduced dramatically with economies of scale.
According to the US Department of Energy (DOE)

2013 Fuel Cell Technologies Market Report (2014a),
the global market for fuel cells grew by almost
400% between 2008 and 2013, with more than
170MW of fuel cell capacity added in 2013 alone
(Figure 10). Currently, more than 80% of fuel cells
are used in stationary applications, such as cogeneration, back-up and remote power systems.
While the UnitedStates ranks first for fuel cell power
capacity, Japan ranks first for delivered systems due
to the successful upscaling of the Ene-Farm micro
co-generation power system.
Although fuel cells saw remarkable development
over the last decade, high investment costs and
relatively limited lifetimes remain the greatest
barriers to their wider application. Investment
costs greatly depend on manufacturing cost, and
could be significantly reduced with economies of
scales. According to the USDOE (USDOE, 2012;
personal contact with US Department of Energy),
PEMFC systems for FCEVs show the highest cost
reduction potential at high production volumes,

and are targeted to ultimately reach costs of
around USD30perkW (Figure 11), which would be
equivalent to ICE engines.
Investment costs for stationary fuel cell systems
are predicted to drop much more slowly, primarily
due to the focus on higher efficiencies and longer
life times. The target cost set by the USDOE
for the 2020 time frame amounts to between
USD1500perkW and USD2000perkW for
medium-sized fuel cell co-generation systems
(USDOE, 2011).
Hydrogen gas turbines

While gas turbines adapted to burn gases with high
hydrogen content (up to 45%) are commercially
available, the same cannot be said for gas
turbines capable of burning pure hydrogen. While
technological modifications would be moderate,
there is currently little demand for such equipment.
31
In the future, gas turbines able to burn very high

shares of hydrogen will be needed for power
generation based on the use of fossil fuels and
pre-combustion CCS, e.g. in integrated gasification
combined cycle (IGCC) power plants with CCS.

This application is currently driving RD&D efforts
in gas turbines able to burn gases with very high
hydrogen content.
Table 10: C
urrent performance of key hydrogen conversion,
T&D and storage technologies
Power or
capacity
Efficiency *
Initial
investment cost
Alkaline FC
Up to 250 kW
~50% (HHV)
PEMFC
stationary
0.5-400 kW
PEMFC mobile
Application
Life time
Maturity
USD 200-700/kW
5 000-8 000
hours
Early market
32%-49% (HHV)
USD 3 000-4 000/kW
~60 000
hours
Early market
80-100 kW
Up to 60% (HHV)
USD ~500/kW
SOFC
Up to 200 kW
50%-70% (HHV)
USD 3 000-4 000/kW
Up to 90 000
hours
Demonstration
PAFC
Up to 11 MW
30%-40% (HHV)
USD 4 000-5 000/kW
30 00060 000 hours
Mature
MCFC
KW to several
MW
More than
60% (HHV)
USD 4 000-6 000/kW
20 000Early market
30 000 hours
Compressor,
18 MPa
88%-95%
USD ~70 /kWH2
20 years
Mature
Compressor,
70 MPa
80%-91%
USD 200-400/kWH2
20 years
Early market
Liquefier
15-80 MW
~70%
USD 900-2 000/kW
30 years
Mature
FCEV on-board
storage tank,
70 MPa
5 to 6 kg H2
Almost 100% (without

compression)
USD 33-17/kWh (10000

and 500 000 units
produced per year)
15 years
Early market
Pressurised
tank
0.1-10 MWh
Almost 100% (without

compression)
USD 6 000-10 000/MWh
20 years
Mature
Liquid storage
0.1-100 GWh
Boil-off stream: 0.3%

loss per day
USD 800-10 000/MWh
20 years
Mature
95%,
incl. compression
Rural: USD 300 0001.2 million/km Urban:

USD 700 000-1.5 million
/km (dependent on
diameter)
40 years
Mature
Pipeline
<5 000 hours Early market

Sources: IEA data; Blum et al. (2014), Overview on the Jlich SOFC development status, 11th European SOFC & SOE Forum, Lucerne;
Decourt et al. (2014), Hydrogen-Based Energy Conversion, More Than Storage: System Flexibility; ETSAP (2014), Hydrogen Production and
Distribution; IEA AFC IA (2015), International Status of Molten Carbonate Fuel Cells Technology; NREL (2009a), Scenario development
and analysis of hydrogen as a large-scale energy storage medium, RMEL Meeting; NREL (2010), Molten Carbonate and Phosphoric Acid
Stationary Fuel Cells: Overview and Gap Analysis; NREL (2009b), Scenario Development and Analysis of Hydrogen as a Large-Scale Energy
Storage Medium; Saur (2008), Wind-To-Hydrogen Project: Electrolyzer Capital Cost Study; Schaber, Steinke and Hamacher (2013), Managing
temporary oversupply from renewables efficiently: electricity storage versus energy sector coupling in Germany; Stolzenburg et al.
(2014), Integration von Wind-Wasserstoff-Systemen in das Energiesystem Abschlussbericht; US DOE (2014b), Hydrogen and Fuel Cells
Program Record; US DOE (2014d), DOE Fuel Cell Technologies Office Record Fuel Cell System Costs; US DOE (2013), Fuel Cell Technology
Office Record - Onboard Type IV Compressed Hydrogen Storage Systems Current Performance and Cost.
32
Gas turbines able to react rapidly to changes in

gas quality, especially with respect to hydrogen
content, are necessary if blending hydrogen in
the natural gas grid (power-to-gas) is to become a
means of integrating otherwise-curtailed renewable
power into the power sector.
Compressors
Compressors are a key technology for hydrogen
storage. Hydrogen pressure levels range from 2MPa
to 18MPa for underground storage, over 35MPa
to 50MPa for gaseous truck transport and up to
70MPa for on-board storage in FCEVs. A recent
study from the USNational Renewable Energy
Laboratory (NREL) concluded that very sparse
data are available on compression technology at
very high pressures (e.g. needed for FCEV onboard storage), with energy demand necessary
for compression varying by a factor of ten among
technologies (NREL, 2014). This is largely due to the
fact that to date such high pressure compressors
are produced in small numbers, as only very little
demand exists.
Hydrogen storage in tanks and

solid structures
Mature options for storage of hydrogen in vessels
comprise pressurised and cryogenic tanks,
providing hydrogen storage capacities of between
100kilowatt hours (kWh) (pressurised tanks) and
100GWh (cryogenic storage). While pressurised
tanks have high costs due to their limited energy
density, cryogenic tanks provide limited storage
time due to the boil-off stream losses, necessary
to maintain acceptable pressure levels. An
intermediary solution between pressurised and
cryogenic hydrogen storage is cryo-compressed
hydrogen. In this case, liquefied hydrogen is filled
to the tank, but the pressures levels until hydrogen
needs to be flared are much higher (up to 35MPa)
compared to cryogenic storage (around 2 to 4MPa).
This allows cryo-compressed hydrogen to be stored
for longer time periods.
Storing hydrogen in metal hydrides or carbon
nano-structures are promising technology options
for achieving high volumetric densities. While metal
hydrides are already in the demonstration phase,
fundamental research is still needed to better
understand the potential of carbon nano-structures.
33

The ETP 2DS is set up to attain a carbon emission
trajectory that limits global warming to 2C. By
2050, total global energy-related carbon emissions
need to more than halve compared to current
levels. All energy sectors need to contribute if this
ambitious target is to be achieved (Figure 12).
Energy supply, including the power generation
and fuel transformation sectors (termed Other
transformation in Figure 12) will need to contribute
almost half of the emission reductions.
Low-carbon energy systems largely rely on the deep
decarbonisation of the power sector. The increased
deployment of renewable energy, such as wind,
solar, biomass and hydropower, is a key element in
the supply of low-carbon electricity by 2050, the

global share of renewable electricity in the power
sector is as high as 63% in the 2DS.6
The remaining half of the emission reduction
necessary to achieve a 2C trajectory will need to
come from the energy demand sectors, namely
transport, buildings and industry. This largely
depends on the deployment of highly efficient enduse technologies, switching to low-carbon fuels such
as hydrogen or advanced biofuels, or avoiding the
use of energy through reduced activity levels e.g. in
the transport sector.
6. Renewable energy includes VRE such as wind, solar photovoltaic

and ocean energy, plus renewable energy which is not classified
as variable, such as hydropower, biomass, geothermal and
concentrated solar power (CSP).
Figure 12: Energy-related carbon emission reductions by sector in the ETP 2DS
60
Transport 20%
50
Industry 21%
GtCO2
40
Buildings 12%
30
Other transformation 8%
20
Power generation 39%
10
0
2012
2020
2030
2040
2050
Note: GtCO2 = gigatonnes of carbon dioxide.
KEY POINT: All energy sectors need to contribute to achieve the ETP 2DS.
The following sections outline how the intensified

deployment of hydrogen technologies, essentially
the Vision, could contribute to achieving the
2DS. Detailed modelling results are provided
for transport, based on a variant of the 2DS: the
2DShighH2. As the analysis of hydrogen T&D
and retail infrastructure requires country-specific
evaluation, detailed results are only provided
for the UnitedStates, Japan and EU4 (France,
Italy, Germany and the UnitedKingdom). A
benchmarking approach is used in the power
and industrial sectors, while a more qualitative
discussion covers the vision for hydrogen
applications in buildings.
34
Transport
Although global transport activity is estimated to
double between now and 2050 under a businessas-usual scenario, transport-related carbon
emissions are halved compared with 2012 in the
2DS, thus contributing about 20% to total energyrelated emission reductions (Figure 12). In 2012,
road transport accounted for 75% of all transport
emissions. It will therefore have to contribute the
largest share to total transport sector emission
reductions in the future.
Based on the ETP Avoid-Shift-Improve concept,

carbon emissions can be reduced by: 1) avoiding
travel, e.g. due to better urban planning and
a significantly increased share of teleworking;
2) shifting travel to more efficient modes such
as public passenger transport and rail freight;
and 3) improving transport technologies. The
improve option includes increasing the efficiency
of conventional technologies, the rapid uptake
of alternative vehicles such as BEVs, PHEVs and

FCEVs, as well as switching to advanced biofuels, in
particular for long-distance transport modes, such
as long-haul road freight, air and shipping.
Box 5 provides an explanation of the ETP scenarios,
and introduces the 2DShighH2, which is then
explored in further detail in the section below.
Box 5: ETP scenarios and the hydrogen roadmap variant FCEV roll-out scenario
The ETP 2DS describes how technologies and
energy-use patterns across all energy sectors
may be transformed by 2050 to give a 50%
chance of limiting average global temperature
increase to 2C. It sets a target of cutting energyrelated CO2 emissions by more than half by 2050
(compared with 2012). The 2DS acknowledges
that transforming the energy sector is vital
but not the sufficient solution; the goal can
only be achieved if CO2 and GHG emissions in
non-energy sectors are also reduced. The 2DS
is broadly consistent with the World Energy
Outlook 450 Scenario through to 2040.
The model used for the analysis of the power
and fuel transformation sectors is a bottom-up
TIMES* model that uses cost optimisation to
identify least-cost mixes of technologies and
fuels to meet energy demand, given constraints
such as the availability of natural resources. The
TIMES energy supply model, which has been
used in many analyses of the global energy
sector, is supplemented by detailed demandside simulation models for all major end-uses in
the industry, buildings and transport sectors.
The IEA ETP 6DS is largely an extension of
current trends. By 2050, global energy use
increases by 75% (compared with 2015) and
total GHG emissions rise by almost 60%. In
the absence of efforts to stabilise atmospheric
concentrations of GHGs, the average global
temperature is projected to rise to up to 6C in
the long term. The 6DS is broadly consistent
with the World Energy Outlook Current Policy
Scenario through to 2040.
* TIMES stands for The Integrated MARKAL-EFOM System.
For the analysis of the transport sector, this

roadmap builds on a variant of the ETP 2DS
the ETP 2DS high H2. This scenario explores the
effects on energy use, CO2 emissions and costs
if hydrogen enters the transport sector earlier
and to a much greater extent than in the ETP
2DS, while delivering comparable emission
reductions. The 2DS high H2 follows what if
logic, assuming twice the amount of FCEVs
in both passenger (25% of the PLDV stock by
2050) and freight road transport (10% of the
LCV, MFT and HFT stock by 2050) compared to
the ETP 2DS. While the share of BEVs stays the
same, the share of PHEVs is reduced in the ETP
2DS high H2 compared to the ETP 2DS.
The rationale for investigating such a variant
is based on the fact that a great deal of
uncertainty surrounds technology choice
in road transport, mainly due to the lack of
maturity and commercialisation of key lowcarbon transport options, as well as stilllimited experience with respect to consumer
acceptance. This, as a consequence, results in
increased uncertainty about their long-term
cost as carbon mitigation measures. In ETP 2012
(IEA, 2012), a sensitivity analysis was conducted
to better understand the role of hydrogen
technologies in transport. Three cases, the ETP
2DS, the 2DS high H2 and the 2DS no H2 were
compared with respect to carbon mitigation
potential and cost. In a mature market, different
mitigation options for the same mobility service
hybrids and PHEVs together with high shares
of sustainable and low-carbon biofuels versus
FCEVs in combination with low-carbon footprint
hydrogen showed comparable total costs and
mitigation potential. However, the transition
towards these different transport futures hinges
on overcoming different barriers.
35
Box 5: ETP s cenarios and the hydrogen roadmap variant FCEV roll-out scenario
(continued)
With the large-scale deployment of hydrogen
technologies in transport, the economic barriers
linked to the establishment of the hydrogen
infrastructure are reduced if combined with
rapid technology adoption, higher FCEV market
penetration and thus higher hydrogen demand.
FCEV roll-out scenario

The PLDV stock within Europe, Japan and the
UnitedStates is already close to the saturation
point (Figure 13). Similar to the ETP 2DS, under the
2DShighH2 total vehicle ownership is affected by
avoiding transport and shifting demand to more
efficient public transport modes.
By building on the most optimistic scenario,

this roadmap provides insights to an ambitious
and yet feasible scenario aiming to minimise
the need for subsidies to achieve parity of cost
between FCEVs and efficient conventional
vehicle technology.
The technology profiles of the PLDV fleet differ

across the regions, in particular noting the large
share of diesel passenger cars in Europe. In the
future, the share of conventional ICE vehicles and
hybrids without the option to plug into the power
grid will need to drop to around 30% of the vehicle
fleet, in order to attain the 2DS.
Figure 13: P
LDV stock by technology for the UnitedStates,
EU4 and Japan in the 2DShighH2
United States
PLDV stock, millions
250
160
Japan
EU 4
70
140
200
60
120
50
100
150
40
80
100
30
60
20
40
50
10
20
0
2010
2020
2030
2040
2050
Conventional ICE gasoline
Hybrid gasoline
Plug-in hybrid gasoline
Plug-in hybrid diesel
0
2010
2020
2030
2040
Conventional ICE diesel
BEV
0
2010
2050
Hybrid diesel
FCEV
2020
2030
2040
2050
Conventional ICE CNG/LPG
Note: LPG = liquefied petroleum gas.
KEY POINT: While in all regions the share of conventional vehicles drops below 10% by 2050,
the technology mix remains region-specific.
Compared to the ETP 2DS, the higher share of FCEVs

in the 2DShighH2 displaces some of the plug-in
hybrid vehicles. While this ensures a similar emission
trajectory, the need for biofuels is significantly
reduced in the 2DShighH2.
The FCEV sales scenario necessary to reach the
PLDV stock shown in Figure 13 is very ambitious it
assumes that by 2020 around 30000 FCEVs will
have been sold in the UnitedStates, EU4 and Japan.
Cumulative sales reach about 8millionFCEVs by
36
2030. By 2050, the share of FCEVs in total passenger

car sales is around 30%. Since FCEV production
costs strongly depend on annual production rates,
the fast ramp-up of FCEV sales is a prerequisite
for rapidly decreasing FCEV cost. An overview
of costs of PLDVs by technology as used in the
model for the UnitedStates is provided in Table 11,
detailed techno-economic parameters for FCEVs are
provided in Table 12.
Box 6: The economics of renewable hydrogen
250
LCOH2 as funcon of electricity and carbon price
200
150
100
50
0
0
20
45
70
95
120
Electrcity price USD/MWh
Carbon price USD/tCO2
PEM EL
NG CCS low
NG CCS high
NG medium
145
175
Levelised costs of H 2 USD/MWh
Levelised costs of H2 USD/MWh
Figure 14: C
ost of hydrogen as a function of electricity price
and annual load factor
250
LCOH2 as funcon of annual ulisaon
200
150
100
50
0
0%
20%
40%
60%
80%
100%
Annual ulisaon factor

PEM EL sensivity
NG CCS medium
NG low
NG high
PEM EL 100 USD/MWh

PEM EL 20 USD/MWh
NG CCS high
PEM EL 60 USD/MWh
NG CCS medium
NG CCS low
Notes: PEM EL = proton exchange membrane electrolyser; LCOE = levelised cost of energy; NG = natural gas; for the lefthand graph, annual loads of 85% are assumed for all technology options, and the dashed line marks the sensitivity of LCOE
for hydrogen from a PEM electrolyser with a 30% variation in cost and a 10% variation in efficiency.
For hydrogen from natural gas, the terms low, medium and high denote: low natural gas price: USD20perMWh, no T&D;
medium natural gas price: USD40perMWh, T&D: USD25perMWh of H2; high natural gas price: USD60perMWh, T&D:
USD50perMWh of H2.
KEY POINT: Low-carbon electrolytic hydrogen requires low-cost renewable electricity

and a combination of higher natural gas and carbon prices to be cost competitive.
The adoption of renewable hydrogen versus
the use of fossil-derived hydrogen (with or
without CCS) strongly depends on its economic
competiveness. The relationships between
natural gas price, electricity price, annual
full-load hours, carbon price and the resulting
cost of hydrogen are illustrated in Figure 14.
Even under optimistic assumptions with regard
to the techno-economic parameters of the
electrolyser, electrolytic hydrogen remains
considerably more expensive than hydrogen
from natural gas reforming, unless very lowcost renewable electricity is available and
carbon or natural gas prices are high.
This is especially true as a combination of very
low costs for electricity from VRE together with
annual full-load factors above 30% is unlikely.
While the results in the left-hand graph in
Figure 14 are based on annual utilisation factors
of 85%, the right-hand graph demonstrates the
link between electricity prices, annual utilisation
factors and hydrogen costs.
Natural gas SMR in combination with CCS

appears to be an attractive option for hydrogen
generation, if the carbon price is above
USD50per tonne of CO2. At low natural gas
prices, renewable hydrogen would only be cost
competitive if low-cost, low-carbon electricity
was available for more than 80% of the hours of
the year.
However, looking purely at hydrogen generation
costs is not enough costs for hydrogen T&D
need to be taken into account to evaluate the
competitiveness of renewable hydrogen. In
case of large-scale natural gas SMR, hydrogen
production takes place at centralised production
facilities and transmission distances to the point
of hydrogen use can be long, leading to high
transmission costs. By contrast, the decentralised
production of hydrogen via electrolysis can make
hydrogen T&D obsolete.
At moderate daily hydrogen demand in the
order of several tonnes, transmission costs of
around USD50perMWh occur if hydrogen
needs to be transported over a longer distance
of 100km; by comparison, if a shorter
37
Box 6: The economics of renewable hydrogen (continued)

distance of 50km is seen, T&D amounts to
USD25perMWh (Yang & Ogden, 2007).
Adding the transmission costs to the generation
costs for centrally produced hydrogen using
natural gas at USD40perMWh* together with
SMR and CCS (NG CCS medium), renewable
hydrogen could be cost competitive if
* For comparison, current natural gas prices account for
USD13perMWh in the UnitedStates, USD37perMWh in
the EuropeanUnion and USD56perMWh in Japan.
produced from electricity at prices of up to

USD20perMWh and at lower annual utilisation
factors of around 15%. Alternatively, electricity
prices of up to USD60perMWh at high annual
utilisation factors of around 80% would also
result in cost competitiveness with centrally
produced low-carbon footprint hydrogen (NG
CCS medium).
Table 11: C
ost of PLDVs by technology as computed
in the model for the UnitedStates
Today
28600
Conventional ICE diesel
29300
Hybrid gasoline
30000
32400
BEV (150 km)
35400
FCEV
60000
2030
30900
2050
32300
Unit
USD
31700
33100
USD
31800
33200
USD
33200
34400
USD
32800
34000
USD
33600
33400
USD
Note: In line with results from the National Academy of Science report on Transitions to Alternative Vehicles and Fuels (National
Research Council, 2013,) FCEVs become less expensive than plug-in hybrids by 2050. Similar tables showing the costs of PLDVs as
computed by the model for Europe and Japan can be found in the Technology Annex.
Table 12: T
echno-economic parameters of FCEVs as computed
in the model for the UnitedStates
Today
60000
2030
33600
2050
33400
Unit
USD
23100
30200
4300
600
1800
24100
4300
3100
460
1600
25600
3200
2800
260
1400
USD
USD
USD
USD
USD
Fuel cell system (80 kW)

H2 tank (6.5 kg H2)
Battery (1.3 kWh)
380
20
460
54
14
350
40
13
200
USD/kW
USD/kWh
USD/kW
Tested fuel economy

Life time
1.0
12
0.8
12
0.6
12
KgH2/100km
Years
FCEV costs
Thereof
Specific costs
Glider*
Fuel cell system**
H2 tank**
Battery**
Electric motor and power control**
Other parameters
Note: The USD DOE Fuel Cell Technology Office Record 13010 suggests total system costs of the 70MPa hydrogen tank of USD33perkWh
at annual production rates of 10000 vehicles, dropping to about USD17perkWh at annual production rates of 10000 vehicles (USDOE,
2013). A tested fuel economy of 0.8kgH2per100km has been reported for the Toyota Mirai (Toyota, 2015a). The assumed tested fuel
economy for todays FCEVs in the UnitedStates is higher based on the assumption that PLDVs are generally larger in the UnitedStates
compared to Japan. They are in line with the results provided in the NREL FCEV demonstration project report (NREL, 2012a).
* future cost increase is due to light-weighting, improved aerodynamics, low resistance tyres and high efficient auxiliary devices.
** future costs are based on learning curves with learning rates of 10% (H 2 tank), 15% (electric motor, power control, battery) and 20%
(fuel cell system) per doubling of cumulative deployment.
38
Carbon footprint of FCEVs

The average on-road WTW CO2 emissions of PLDVs
in todays vehicle stock vary between almost
300gCO2perkm (conventional gasoline ICE
cars in the UnitedStates) and a little more than
75gCO2perkm (BEVs in Japan), depending on the
vehicle power train and the region (see Figure 15).
Although FCEVs and BEVs have no direct CO2
emissions, their WTW carbon footprint is
considerable accounting for emissions linked
to hydrogen and power generation. Based on
hydrogen from both natural gas SMR without
CCS and grid electricity (see hydrogen generation
scenarios Box 7), the CO2 mitigation effect of FCEVs
is moderate when compared to conventional ICE
technology and hybrids, depending on the region.
BEVs and, plug-in hybrids already have lower
WTW emissions.
The specific emissions of FCEVs decline with the

increasing decarbonisation of hydrogen supply, and
finally fall below those of plug-in-hybrids. By 2050,
FCEVs allow for very low-carbon, long-distance
driving. At the same time, the use of FCEVs eases
pressure on biofuel production. Although plug-in
hybrids in combination with high blend shares of
advanced biofuels in the gasoline mix (up to 40%)
achieve comparable reductions in CO2 emissions,
they are dependent on the supply of sustainable
biofuels. Since in the 2DS large amounts of high
energy density, liquid biofuels are already needed
for the decarbonisation of air and water transport,
the reduced demand for sustainable biofuels in the
road transport sector can enable a generally more
sustainable transport pathway, especially in the very
long term after 2050 (IEA, 2012).
Figure 15: S
pecific PLDV stock on-road WTW emissions by technology
for the UnitedStates, EU4 and Japan in the 2DShighH2
WTW gCO2 per km
300
United States
300
EU 4
300
200
200
200
100
100
100
2015
2030
2050
2015
Hybrid gasoline
2030
2050
Japan
2015
BEV
2030
2050
FCEV
Note: Stock on-road WTW emissions include the upstream emissions from liquid fuel production as well as power and hydrogen
generation. The fuel economy of the vehicle stock is based on the fuel economy of new vehicles sold, assumptions on average age and
a gap factor of approximately 20%, accounting for the difference between test-cycle fuel economy and on-road fuel economy.
KEY POINT: While FCEVs currently offer moderate WTW emission reductions compared to conventional
PLDVs, they can enable low-emission, long-distance individual motorised transport in the longer term.
Hydrogen T&D
and refuelling infrastructure
The configuration of hydrogen T&D and retail
infrastructure is determined by many parameters,
including: hydrogen demand; the distance to
the hydrogen production site; the density of the
urban environment; and assumptions on the
required proximity of one station to the next for the

consumer. As previously discussed, whether or not
hydrogen is generated in a centralised manner also
has significant impacts on T&D. As hydrogen T&D
is costly and strongly depends on the transmission
distance, more robust T&D cost estimates can
only realistically be provided in knowledge of the
geographical parameters.
39
Figure 16: S
cheme of hydrogen T&D and retail infrastructure
as represented within the model
Centralised
hydrogen
production
Liquid or gaseous
trucking/pipeline
50 to 150 km
Liquid or gaseous trucking/pipeline
High capacity
terminal
50 km
Sub-terminal
Decentralised
hydrogen
Hydrogen
refueling
station
Liquid or gaseous
trucking/pipeline
Decentralised
hydrogen
Hydrogen
refueling
station
Hydrogen
refueling
station
Small City
Liquid or gaseous
trucking/pipeline
Hydrogen
refueling
station
Big City
KEY POINT: Different hydrogen generation, T&D and retail pathways will develop over time.
In the underlying model scenario, the centralised supply of small cities happens via terminals
based in large urban areas.
Hydrogen station roll-out

Modelling allows for a better understanding of the
interaction between the key variables in the delivery
of hydrogen T&D and retail infrastructure for FCEVs.
A simple hydrogen T&D model has been introduced
to the IEA Mobility Model. It distinguishes between
hydrogen demand in large cities (~500000
inhabitants) and hydrogen demand in small urban
areas (~25000 inhabitants), representing the
size of typical settlements in the regions under
discussion. As hydrogen transmission distance
has a high impact on T&D costs, assumptions on
40
average transmission distance between sites of

production and demand, as well as the density of
the retail station network, have a strong influence
on hydrogen T&D costs. A schematic drawing
illustrating some of the variables is provided in
Figure 16.
Refuelling stations of two different sizes are
represented in the model: daily refuelling capacities
of 500kg and 1800kg. Construction of small or
large hydrogen stations depends on the FCEV fleet
and the respective daily hydrogen demand.
Hydrogen can be delivered by truck (either in

gaseous or liquefied form), or as a gas using
hydrogen pipelines. Which approach to adopt for
both long-distance hydrogen transmission as well as
for intra-urban distribution is determined based on
the optimal hydrogen T&D strategies provided in
Yang and Ogdens paper (2007).
In the initial phase, hydrogen refuelling
infrastructure is only available in large urban
agglomerations and on some connecting routes.
Based on regional differences in population density,

assumptions on average speed and maximum
acceptable time to the next hydrogen station, a
basic station network is established. After 2030, in
order to ramp up FCEV sales, the network needs to
be expanded to average-sized cities. By 2040, in
the 2DShighH2 all cities need to have at least one
hydrogen station.
Box 7: Spotlight on hydrogen generation

Within the 2DS high H2, hydrogen is
produced from a broad variety of energy
sources, depending on region-specific
resource endowment, and subject to the
carbon emission constraint to meet a 2C
target. Unlike the power generation mix,
which is a result of the TIMES optimisation
tool, hydrogen generation is based on a
simulation approach, and the different
hydrogen generation pathways are defined
exogenously. In the model, hydrogen can
be produced from natural gas via largescale SMR with or without CCS, from coal
via coal gasification and reformation with
CCS, from biomass and from electricity.
For hydrogen production from electricity
via electrolysis, either grid electricity* or
low-carbon, low-cost surplus** electricity
is used. For all generation pathways,
levelised cost of hydrogen generation is
calculated based on economic parameters
such as investment costs, fuel costs, carbon
price, operation and maintenance costs
and interest rate, as well as technological
parameters such as conversion efficiencies,
lifetimes and annual utilisation factors.
Various other hydrogen generation
pathways e.g. reformation of biogas or
hydrogen production from waste water do
exist but are not modelled explicitely.
Surplus electricity is currently curtailed if there

is no demand and supply cannot be varied,
leading to a market value of zero. In the future,
otherwise-curtailed electricity will also have a
lower price than its levelised cost of production.
Estimating the amount of curtailed electricity in
low-carbon power systems is a prerequisite to
quantify the potential amount of cost-effective,
renewable hydrogen. The amount of curtailed
electricity as a function of the share of VRE in
the power system has been recently published
in a study by NREL (2012b). According to their
results, annual curtailment levels of renewable
power in the United States reach between 60
TWh and 150 TWh (4% to 7% of the generated
renewable power per year), at renewable power
shares between 60% and 90%. In theory, this
would be sufficient to supply a fleet of around
6 million to 16 million FCEVs with renewable
hydrogen. Other studies, (e.g. Mansilla, 2013;
Jorgensen, 2008) document evidence of
significant annual hours with electricity spot
market prices of below USD 20 per MWh in
countries with high VRE penetration.
The hydrogen generation pathways shown in
Figure 17 are defined to meet the 2DS emission
target at lowest cost and include carbon
prices for emissions occurring during the fuel
production process, which gradually increase
up to USD 150 per tonne of CO2 by 2050.
* Within 2DS high H2 the electricity mix of the 2DS published

in ETP 2014 (IEA, 2014a) is used. Until 2050, power generation
is almost completely decarbonised to meet the 2C target.
** Surplus electricity refers to electricity from VRE, which
cannot be fed into the power grid due to either temporal or
geographical mismatch between electricity generation and
demand.
41
Box 7: Spotlight on hydrogen generation (continued)

Figure 17: H
ydrogen generation by technology for the 2DShighH2
in the UnitedStates, EU4 and Japan
TWh
Japan
EU 4
United States
450
400
350
300
250
200
150
100
50
0
2010
140
120
100
80
60
40
20
2020
Natural gas
2030
2040
Natural gas and CCS
0
2010
2050
Coal and CCS
2020
2030
2040
Biomass gasification
2050
45
40
35
30
25
20
15
10
5
0
2010
Average mix electrcity
2020
2030
2040
2050
Low cost renewable electricity
KEY POINT: Hydrogen supply depends on regionally different resource endowments.

During the early years, most of the hydrogen is
supplied using natural gas SMR without CCS.
After 2030, no new SMR capacity without CCS
is added, since SMR with CCS*** is becoming
cost competitive due to CO2 prices of around
USD 90 per tonne. Hydrogen from renewable
electricity is only cost effective if low-cost,
surplus electricity is used. Grid electricity at
future retail prices (2050) of USD 115 (United

States) to USD 137 (EU 4) per MWh is assumed
to be cost-prohibitive, even if T&D costs are
zero (Figure 18). It is estimated that low-cost,
surplus renewable power would be sufficient
to supply between 12% (Japan) and 30% (EU 4)
of the hydrogen used in transport by 2050.****
Hydrogen production from biomass is assumed
to play a minor role in all three regions.
*** For CCS technologies, capture rates of 80% are assumed.

Furthermore, conservative cost estimates of USD 20 per
tonne of CO2 for transport and storage are included.
**** It is assumed that around 3% to 7% of annual renewable

power generation is available at prices of around USD 20
to USD 30 per MWh for 1 370 to 2 140 hours of the year,
depending on the region.
EU 4
100
Japan
Biomass gasification
Electrolysis excess
power
Electroysis grid
power
0
Coal CCS
Electroysis grid power
United States
NG SMR CCS
100
2050
200
NG SMR
200
Today
NG SMR
Figure 18: Hydrogen production costs without T&D for the 2DShighH2
KEY POINT: Hydrogen produced from grid electricity is costly compared to alternative
generation pathways. For cost-effective renewable hydrogen, the availability of low-cost,
surplus renewable electricity is a prerequisite.*****
***** All underlying technoeconomic assumptions can be found in Table 15
42
To cost effectively meet future hydrogen

demand, an important share of generation
is based on fossil fuels in combination with
CCS.****** Alternative scenarios envisaging
higher shares of hydrogen from renewable
electricity are feasible, especially if the use of
CCS is constrained by political choices or a lack
of available CO2 storage resources, although
these alternatives are more costly. As hydrogen
produced from grid electricity is significantly
more expensive than hydrogen from SMR or

from low-cost, surplus electricity, this will affect
the cost of hydrogen at the station. Hydrogen
demand from the transport sector accounts
for between 1% (EU 4 and Japan) and 3%
(United States) of total final energy demand
and between 4% (Japan) and 10% (United
States) of total electricity demand in 2050.
Significantly increasing the share of hydrogen
from renewable electricity in the generation
mix would require substantial additions to
renewable power capacity.
****** In Japan, hydrogen from natural gas or coal with CCS

is assumed to be imported either as liquefied hydrogen or
in chemically bound form. Transport costs are taken from:
Inoue (2012).
A station roll-out scenario, which corresponds to the

assumed FCEV penetration, is shown in Figure19. As
annual mileage in the UnitedStates is much higher
than in Europe and Japan, an earlier and more
widespread demand for large hydrogen stations is
needed. More detailed results on T&D infrastructure
requirements can be found in the Technology
Annex for the regions under discussion.
Although the approach used is based on a
simplistic representation of hydrogen refuelling
infrastructure, it reveals some interesting insights.
While the provision of affordable hydrogen in
densely populated large urban areas can be reached
over the course of tenyears, it rapidly becomes
costly when expanding the network to average-
sized cities, as the number of hydrogen stations

necessary to provide national and regional coverage
quickly jumps to several thousands.
While within this exercise only two station sizes
(500kgperday and 1800kgperday) are taken
into account, much smaller stations might be
needed in the initial phase to achieve widespread
coverage. A 100kgperday station would be
enough to refill a fleet of around 200 FCEVs
over the year,7 and could fulfil basic needs while
avoiding excessive under-utilisation.
7. A ssuming 10000kmperyear, 1.1kg hydrogen per 100km, and

an annual load factor of 60% for the station.
2015
2020
2025
2030
2035
2040
EU 4
Japan
United States
EU 4
2045
Japan
United States
EU 4
Japan
United States
Japan
EU 4
United States
0
Japan
5 000
EU 4
100
United States
500 kg
10 000
EU 4
15 000
200
Japan
300
United States
1800 kg
EU 4
20 000
Japan
25 000
400
Japan
30 000
500
United States
35 000
600
EU 4
700
United States
Number of staons
Figure 19: H
ydrogen stations for the 2DShighH2 in the UnitedStates,
EU 4 and Japan
2050
Note: By the end of 2015 already 100 hydrogen stations are planned to be built in Japan.
KEY POINT: Due to higher mileages and larger cars,

larger hydrogen stations are needed in the UnitedStates.
43
The necessary hydrogen generation, T&D and retail

infrastructure requires significant investment in the
order of tens to hundreds of billions of dollars. As
for any other fuel, these investments are recovered
through the fuel price when selling hydrogen to
the consumer.
Within the 2DShighH2, the total cumulative
investment in hydrogen generation, T&D and
retail infrastructure up to 2050 amounts to
USD140billion, USD50billion and USD26billion
for the UnitedStates, EU4 and Japan respectively.
A more detailed breakdown of costs is provided in
the Technology Annex. For each of the 150million
FCEVs sold between now and 2050 in the United
States, EU4 and Japan, between USD900 to (Japan)
to USD1900 (United States) needs to be invested in
the built-up of the hydrogen generation, T&D and
retail infrastructure.
Total cost of driving and subsidy requirements
Total cost of driving (TCD, see Box 8) can be used
to estimate the level of subsidy necessary for FCEVs
to achieve parity of cost with a benchmarking
technology. For this purpose, efficient conventional
cars have been selected as the benchmark within
the 2DShighH2.
During the market introduction phase, the TCD for
FCEVs is much higher than for efficient conventional
cars of the same segment (Figure 20). This is due
to higher vehicle investment and higher fuel costs.
While TCD on a pure cost basis reach parity with
efficient conventional cars in the United States by
2040 in the 2DS high H2, TCD of FCEVs will always
stay higher in EU 4 and Japan due to higher costs

for hydrogen production. In order to represent
the full range of costs to the consumer, regional
fuel taxation schemes need to be included when
calculating TCD tax. Currently, petroleum taxes in
the EU4 and Japan are equal to around 100% of the
cost of fuel at the station, while in the UnitedStates
petroleum fuel taxes are far lower. For the purposes
of projection, petroleum tax levels are assumed
to stay constant for the EU4 and Japan, while a
moderate 30% petroleum tax is assumed for the
UnitedStates.
Differentiated fuel taxation provides a mean to
reach parity of cost of TCD tax with conventional cars
as soon as possible. For that purpose, in contrast
to petroleum fuels, hydrogen needs to remain
untaxed until cost parity is achieved. If hydrogen
were exempted from fuel taxation, FCEVs would
become cost competitive with efficient conventional
cars around 2035 in all three regions under the
2DShighH2.
To generate early consumer interest in FCEVs,
the differential in TCD tax with the benchmark
technology before cost parity is achieved could be
covered by direct subsidies, in addition to the fuel
tax exemption. Figure 21 illustrates the relationship
between FCEV vehicle stock (blue line), the absolute
amount of direct subsidy per vehicle necessary to
achieve cost parity with efficient conventional cars
(orange line), and total FCEV government subsidy as
a percentage of total petroleum tax revenue in that
year (green line).
Box 8: Vehicle costs, fuel cost and TCD

TCD is a valuable measure to compare
different vehicle technologies on the basis of
economics. TCD (expressed in USDperkm)
comprises the cost of the vehicle and fuel over
the entire vehicle lifetime, divided by total
driven kilometres.* Fuel costs comprise the
cost of the fuel and costs relating to the T&D
and retail infrastructure. When comparing
vehicle technologies on a purely economic
basis, non-monetary consumer preferences
such as range, technology, image or subjective

notions of reliability are not taken into account.
In the following text, the US example will be
discussed; results for EU4 and Japan can be
found in the Technology Annex.
With a rapid ramp-up of FCEV sales, vehicle
costs drop quickly (Figure 20). Cost parity with
plug-in hybrids is reached by 2030, and by
2040 FCEVs have almost reached the cost of
conventional hybrids.
* With this approach, vehicle costs are depreciated over the

entire vehicle lifetime. A discount rate of zero is applied.
44
60
4.8
50
3.6
40
2.4
30
1.2
20
2010
0.0
2020
2030
FCEV
Hybrid gasoline
Costs of H2 at the staon
2040
2050
Convenonal ICE gasoline

BEV
Sensivity H2 costs
0.5
80
0.4
60
0.3
40
0.2
20
0.1
2010
2020
2030
FCEV
Convenonal ICE gasoline
BEV
FCEV stock millions
FCEV fleet, millions
6.0
Total costs of driving
70
Costs of H2 USD/Lge
Vehicle costs, thousand USD
Figure 20: V
ehicle costs, fuel costs and TCD for FCEVs
in the 2DShighH2 in the UnitedStates
0
2040
2050
FCEV TCD sensivity
Hybrid gasoline
KEY POINT: While FCEV costs drop rapidly as sales ramp up, the cost of hydrogen at the pump
drops much more slowly. Hydrogen costs decline quickly as long as stations are clustered around
early demand centres. When the hydrogen refuelling network is expanded to provide the coverage
necessary to sell millions of FCEVs, hydrogen costs see another increase.
Costs of hydrogen at the station drop much

more slowly, partly due to under-utilisation of
the T&D and retail infrastructure. While the
provision of affordable hydrogen in densely
populated large urban areas can be reached
over the course of ten years, costs increase
significantly when expanding the refuelling
network to average-sized cities, which in this
modelling approach is assumed to take place
after 2025. Network expansion to connect
the average city causes hydrogen costs to
increase again, due to longer T&D distances,
smaller stations and under-utilisation of
the infrastructure.** However, nationwide
coverage of the hydrogen refuelling network is
a prerequisite to bringing millions of FCEVs on
the road.
** The fixed costs of the hydrogen T&D infrastructure are

inversely proportional to the utilisation rate.
When comparing the TCD of different

technology options, the relatively high level
of costs associated with hydrogen delays cost
parity with alternative technology options.
Without any further incentives, the TCD of
FCEVs reaches cost parity with conventional
cars in the UnitedStates by 2040. At 30% lower
hydrogen costs, parity would occur five years
earlier. Assuming 30% higher hydrogen costs,
parity would not occur at all. On the contrary,
due to higher energy costs in Europe and Japan,
FCEVs will not reach cost parity with highefficiency conventional cars without further
incentives, such as differentiated fuel taxation.
When comparing the TCD of FCEVs and BEVs,
it is important to keep in mind that BEVs will
not serve the same mobility service as FCEVs.
BEV users will find that autonomy will be
significantly more limited and recharging time
will be longer. It is hence up to the consumer to
choose between different mobility options at
different costs.
45
While direct subsidy levels per vehicle reach

some USD50000 (United States) to more than
USD 80 000 (Japan) during the very early market
introduction phase, they quickly drop to around
USD5000 per vehicle once the market is more
mature, which is the case around 2025 within the
2DShighH2. Although direct FCEV subsidies reach
annual amounts of up to several billion USD, as a
proportion of annual petroleum fuel tax revenue
they would not exceed 5% in the EU4 and Japan,
and 15% in the UnitedStates.
Based on the scenario results including

differentiated fuel taxation and assuming a rapid
increase in FCEV sales, the market for passenger
FCEVs could be fully sustainable 15 years after
introduction of the first 10000 FCEVs. Once the
subsidy per FCEV (orange line) drops to zero
(i.e.cost parity with conventional cars is reached),
hydrogen could be taxed.
Figure 21: S
ubsidy per FCEV and share of annual subsidy as a percentage
of petroleum fuel tax revenue under the 2DShighH2 in the
UnitedStates, EU 4 and Japan
EU 4
Japan
80
80%
80
80%
80
80%
60
60%
60
60%
60
60%
40
40%
40
40%
40
40%
20
20%
20
20%
20
20%
0%
0%
0
2010
2020
2030
- 20
2040
2050
-20%
FCEV stock
2010
2020
- 20
Subsidy per vehicle sold
2030
2040
0%
2010
2050
2020
2030
2040
-20% - 20
2050
-20%
Share of direct FCEV subsidy on

petroleum tax income
Million vehicles
Thousand USD
United States
Annual share of subsidy on petroleum tax income
KEY POINT: If hydrogen was exempted from fuel taxes and rapid market penetration is assumed, the FCEV
market would be fully sustainable 15years after the introduction of the first 10000FCEVs.
The level of subsidy for FCEVs as a proportion

of petroleum tax revenue reaches its peak when
the FCEV fleet is already in the millions. This
analysis therefore also illustrates that the critical
phase of subsidy for the large-scale introduction
of FCEVs is in the medium rather than the short
term. Cumulative subsidies additional to the fuel
tax exemption of hydrogen to reach cost parity
with efficient gasoline cars account for around
USD59billion in the UnitedStates, USD22billion
in EU4 and USD7.5billion in Japan.
TCD is highly sensitive to the pace of FCEV market
penetration. Were an absence of fuel cell trucks
and light commercial vehicles to be assumed under
this scenario, cumulative subsidies would increase
by around 15%.8 If only 25% of the envisaged
8. The presence of hydrogen trucks significantly affects the
utilisation rates of the hydrogen supply infrastructure, as
hydrogen demand per vehicle is much higher than for passenger
cars. They therefore have an important impact on hydrogen
demand and thus hydrogen costs at the station.
46
passenger FCEVs and no fuel cell trucks at all were

sold by 2050, cost parity with conventional cars
would not be reached, unless consumers would be
willing to pay a 5% to 10% premium per kilometre.
By comparison, the acceptance of a small premium
on TCD of 1% to 2% for FCEVs compared to
conventional cars would reduce the amount of
subsidy significantly.
Carbon mitigation potential
Under the 2DShighH2, the large-scale deployment
of FCEVs in passenger and freight road transport
leads to significant CO2 emission reductions (Figure
22). By 2050, FCEVs contribute around 14% of
the overall annual transport emission reductions
necessary to switch from a 6DS to a 2DS trajectory
in the UnitedStates, EU4 and Japan. Depending
on the region, the contribution of FCEVs to
cumulative total transport CO2 emission reductions
between now and 2050 accounts for between 7%
(UnitedStates) and 10% (Japan). Between now and

2050, almost 3GtCO2 are saved by FCEVs in the
three regions under discussion.
Globally, by 2050 FCEVs could account for almost

1GtCO2 emission reductions per year, assuming a
ten-year delay for FCEV market introduction and
significantly lower growth rates in non-OECD
regions compared to OECD regions.
Figure 22: C
O2 mitigation from FCEVs in transport under the 2DShighH2
in the UnitedStates, EU4 and Japan
EU 4
United States
3 000
1 000
2 500
800
Mt CO2
2 000
Japan
400
300
600
200
1 500
400
1 000
0
2010
100
200
500
2020
2030
2040
2050
0
2010
2020
Other road avoid/shift/improve
2030
2040
2050
Other transport avoid/shift/improve
0
2010
2020
2030
2040
2050
FCEV
KEY POINT: By 2050, the large-scale deployment of FCEVs in the transport sector could account for 14% of
the annual carbon mitigation reductions necessary to switch from a 6DS to a 2DS emission trajectory.
VRE integration
Global electricity demand is estimated to double by
2050 under a business-as-usual scenario, compared
to 2012 (Figure 23). It currently accounts for around
40% of total energy-related carbon emissions (IEA,
2015). The decarbonisation of electricity is pivotal
to achieving the 2DS by 2050 annual powersector carbon emissions need to be reduced by
more than 85% compared to 2012. Lower electricity
demand resulting from more efficient processes
across energy demand sectors accounts for about a
quarter of total power-sector emission reductions.
The remaining three-quarters need to come from
drastic reductions in the carbon profile of electricity
its global average carbon intensity needs to drop
to 40grams of carbon dioxide (gCO2)perkWh by
2050, from around 533gCO2perkWh in 2012.
significantly reduces the need for base load, i.e.

power plants which are designed to run close to
maximum output, almost full load, night and day
during the whole year. Simultaneously, the demand
for mid-merit order and peaking generation tends
to increase, generating power at times of low VRE
output and shutting down when VRE output is
high. The use of energy storage technologies can
effectively reduce the overall need for generation
capacity, by storing power when it is available
and releasing it during times of scarcity, thus
also contributing to better utilisation of existing
generation capacity. However, for a systemfriendly deployment of a mix of renewable energy,
interconnections, demand-side management,
flexible hydropower or thermal generation are
generally less costly options, which warrant being
mobilised first.
A drastic expansion in renewable electricity to

63% of the global power mix also requires the
integration of high shares of VRE up to 42%,
depending on the region. VREs intermittent nature
47
Figure 23: Global electricity generation mix under the 6DS and 2DS
2DS
TWh
6DS
50 000
50 000
40 000
40 000
30 000
30 000
20 000
20 000
10 000
10 000
0
2012
Natural gas
2020
2030
Natural gas with CCS
2040
Oil
Coal
0
2012
2050
Coal with CCS
Nuclear
2020
2030
Biomass and waste
Hydro
2040
Solar
Wind
2050
Other
KEY POINT: The power mix has to change drastically to reach the 2DS while fossil fuels dominate todays
power generation, the share of renewable power needs to increase to 63% globally by 2050.
Deployment of electricity storage

Different electricity storage scenarios have been
investigated in the IEA Technology Roadmap on
Storage (2014b). PHS already provides significant
energy storage capacity of more than 80GW in the
UnitedStates, the European Union, China and India
(Figure 24). While the modelling in the roadmap
does not deal with VRE integration problems, such
as grid congestion or spatial mismatch between
electricity supply and demand, it nonetheless
foresees the deployment of significant levels of
electricity storage for power supply at times of
high demand and low VRE supply. Under the 2DS,
installed storage capacity more than quadruples

and annual electricity output from energy storage
reaches shares of between 3% and 9% of total
VRE power generation in the regions shown
(which account for roughly 60% of global power
generation by 2050).
Under the 2DS, future costs for energy storage are
assumed to be the same as todays costs of PHS.
In the storage breakthrough scenario, investment
costs are assumed to drop by 50%, leading to
the creation of storage capacity six times greater
compared to 2011.
Figure 24: I nstalled electricity storage capacity for selected regions today
and in 2050 under the 2DS and the storage breakthrough scenario
180
160
140
2011
GW
120
100
2050 2DS
80
60
2050 2DS storage

breakthrough
40
20
0
United States
EU 28
China
India
KEY POINT: Under the 2DS, electricity storage accounts for up to 8% of total
installed power capacity by 2050.
48
Hydrogen-based energy storage

Hydrogen-based energy storage systems cover
a broad range of energy storage applications,
with a focus on high power capacity and longer
storage times in the range of hours to weeks, and
even months. Benchmarking hydrogen-based
energy storage systems against mature alternatives
contributes to understanding their potential for
large-scale deployment in the future.
This section compares LCOE of different hydrogenbased storage systems to those of the benchmarking
technology in the respective field of application. In
order to compare energy of the same quality, the
comparison focuses on LCOE of output electricity,
i.e. taking into account the final conversion step to
electricity in case of power-to-gas applications.
Table 13: Power-to-power and power-to-gas systems included in the analysis

Key components
Abbreviation
PEM electrolyser, underground storage, PEMFC
H2PtPPEM/PEM
Alkaline electrolyser, underground storage, PEMFC
H2PtPAlk/PEM
PEM electrolyser, underground storage,

hydrogen open-cycle gas turbine
H2PtPPEM/OCGT
Compressed air storage
CAES
Pumped hydro storage
PHS
PEM electrolyser, NG grid connection, open-cycle gas turbine
H2PtGPEMHENG with OCGT
PEM electrolyser, methaniser, NG grid connection,

open-cycle gas turbine
H2PtGPEMmethane with OCGT
Power-topower
Power-togas
Note: PtP = power to power; PtG = power to gas.
Two different storage applications have been

identified where hydrogen-based systems can play
a vital role inter-seasonal energy storage and daily
arbitrage. Table 13 provides an overview of the
energy storage systems included in this analysis;
the specifications for inter-seasonal storage and

daily arbitrage are provided in Table 14. All technoeconomic parameters of the different technology
options can be found in Table 15.
Table 14: S
ystem specifications for inter-seasonal energy storage
and arbitrage
Unit
Inter-seasonal storage
Arbitrage
Power capacity charging
MW
200
300
Power capacity discharging
MW
500
300
hours
120
Number of cycles per year
274
Annual full-load share
7%
17%
Cost of input electricity
USD/MWh
0-20
0-50
Discharge duration
Note: PtP = power to power; PtG = power to gas.
49
Inter-seasonal energy storage

Inter-seasonal storage allows for temporary shifts
in energy supply over weeks or months. In this
example, the storage systems run fivecharging and
discharging cycles at 500MW power output over a
discharging time of 120hours each. All large-scale
and long-term energy storage systems suffer from
limited annual operation the assumed scheme
leads to 600 full-load hours per year, equivalent
to a 7% annual utilisation factor. LCOE is therefore
high, as the quantity of product sold on which to
generate a return on investment is low.
Figure 25 shows the resulting LCOE for the different
inter-seasonal storage systems listed in Table13
for the years 2030 and 2050. Ranges illustrate the
impact of input electricity cost. All storage options
are benchmarked against an open-cycle gas turbine
(OCGT),9 which would be the alternative for
meeting demand with flexible generation.
For power-to-power applications, hydrogen
systems are most promising for inter-seasonal
energy storage and surplus VRE integration in the
future. They seem to be the only option to deliver
LCOE somewhat close to the benchmark. Due to
the much higher energy density of hydrogen with
underground storage compared to PHS or CAES,
investment costs are shifted from storage to the
conversion technology.
Power-to-gas systems are close to the benchmark
when assuming a low blend share of 5% hydrogen
mixed with natural gas. However, even hydrogen
from zero-cost electricity is expected to cost three
times the natural gas price. Consequently, LCOE
from power-to-gas systems appear unlikely to fall
below the benchmark, unless extremely high CO2
prices of more than USD 400 per tonne of CO2are
assumed.
An advantage of a power-to-gas system that uses a

gas turbine for re-electrification of a natural gashydrogen blend, is that it enables the use of existing
infrastructure (including storage, T&D and reelectrification facilities) whereas alternative energy
storage options rely on systems that need to be built
from scratch. In addition, blending hydrogen into
the gas grid means that the gas turbine can operate
in the electricity market in the usual way. Unlike fuel
cells in power-to-power electricity storage systems,
its annual utilisation factor is not constrained by the
utilisation factor of the electrolyser producing the
hydrogen, which will likely be low if based on lowvalue, surplus renewable electricity.
If gas grids and gas turbines were able to deal
with high and variable hydrogen blend shares
above 20%, power-to-gas systems might also
provide inter-seasonal energy storage by feeding
more hydrogen into the blend at times of low VRE
availability. Because annual average blend shares
could remain at 5% and lower, the higher costs
of the hydrogen in the blend would have only a
marginal impact on average LCOE. In this manner,
existing gas turbines could potentially achieve
higher annual load factors while still meeting a fixed
carbon budget and leveraging the sunk costs of the
existing natural gas infrastructure.
The attractiveness of power-to-gas systems as
a means of integrating high levels of VRE are
economically dependent on declining electrolyser
costs and technically dependent on persistent
imbalances in electricity supply and demand.
From a system perspective, making direct use of
electricity, a high-quality form of energy, should
be a priority wherever practicable. Thus, if future
electricity systems evolve by 2050 to better balance
supply and demand on a daily or seasonal basis, the
window of opportunity for power-to-gas to play a
transitional role in a lowest-cost decarbonisation
pathway may be limited.
9. T
he costs of natural gas are those of the EU in the 2DS
accounting for USD35perMWh in 2030 and USD29perMWh
in 2050. CO2 prices of USD90pertonofCO2 in 2030 and of
USD150pertonofCO2 in 2050 are taken into account. For
the OCGT an annual utilisation factor of 15% is assumed. The
input electricity for the storage systems is assumed to be
100% renewable.
50
Figure 25: L
COE for inter-seasonal energy storage via power-to-power systems
and VRE integration via power-to-gas systems in 2030 and 2050
Levellised costs of electricity
USD/MWh
1,400
1,200
1,000
800
2030
600
2050
400
OCGT benchmark
200
0
H2 PtP
PEM/PEM
H2 PtP
H2 PtP
H2 PtG
PEM HENG
H2 PtG PEM
methan.
PHS
CAES
KEY POINT: Looking ahead, hydrogen-based energy storage systems show the greatest potential
to achieve acceptable LCOE for seasonal storage applications.
Daily arbitrage
Daily arbitrage allows for the shifting of stored
electricity from times of low demand to times of
high demand, taking advantage of the respective
electricity price differential. Operated for daily
arbitrage reasons, future hydrogen-based powerto-power storage systems almost achieve the
performance of pumped hydro or compressed
air storage, if very low-cost electricity is available
(Figure 26). As the overall efficiency of the
hydrogen-based systems is lower than for PHS and
CAES (see Table 15), LCOE10 is much more sensitive
to the cost of electricity. Hydrogen-based storage
systems would only work for arbitrage reasons with
a very high spread of electricity prices between
times of low and high demand. To break even with
the 2DS storage benchmark, input electricity should
cost no more than USD10perMWh by 2030 and
USD25perMWh by 2050, which is unlikely to
happen at the assumed annual utilisation factor for
daily arbitrage.
For hydrogen-based power-to-power storage
systems to achieve LCOE of USD90perMWh, as in
the breakthrough scenario, the cost of investment
attributable to both the electrolyser and the fuel cell
would need to drop to around USD400perMWh,
and efficiencies would need to increase to up to
10.
The costs of natural gas are those of the EuropeanUnion
under the 2DS, accounting for USD35 per MWh in 2030 and
USD29 per MWh in 2050. CO2 prices of USD90 per tonne of
CO2 in 2030 and USD150per tonne ofCO2 in 2050 are taken
into account. The input electricity for the storage systems is
assumed to be 100% renewable.
90% for electrolysers and 60% for fuel cells (HHV).

In addition, the electricity for arbitrage should cost
no more than USD20perMWh. Consequently,
it seems unlikely that hydrogen-based power-topower storage systems will attain the breakthrough
cost target.
Marginal abatement costs

of hydrogen-based energy
storage options
Examining CO2 abatement costs allows the benefits
of using hydrogen from otherwise-curtailed
renewable power to be compared, addressing
power-to-power, power-to-gas or power-to-fuel
systems (Figure 27).11 To ensure comparability,
all systems are attributed the same annual fullload hours and input electricity prices in the same
region. Figure 27 shows cost ranges among the
UnitedStates, EU4 and Japan. Cross-regional
differences are caused by varying natural gas and
electricity prices, as well as different annual fullload hours.12
11.
Power-to-power and power-to-gas systems are benchmarked
against OCGTs fuelled with natural gas and operated at the
same annual full-load hours as the energy storage system.
Power-to-fuel applications are benchmarked against future
high-efficiency gasoline vehicles fuelled with gasoline blend
containing 30% second-generation biofuels.
12. A
ssumed costs for low-value VRE electricity range between

USD20 per MWh in the EU4 and USD30 per MWh in Japan,
while annual full-load hours range between 1370 hours in
Japan and 2130 hours in the EU4.
51
Figure 26: L
COE of different energy storage technologies
for daily arbitrage in 2030 and 2050
Levelised costs of electricity
USD/MWh
300
250
2030
200
2050
150
2DS storage benchmark
100
2DS storage breakthrough
benchmark
50
0
H2 PtP
PEM/PEM
H2 PtP
PHS
H2 PtP
CAES
KEY POINT: Hydrogen-based storage technologies can be competitive at low electricity input prices.
In the long term, power-to-fuel applications offer the

lowest marginal abatement costs for hydrogen-based
VRE integration. Cost reductions due to technological
gains are greater than for other storage applications,
as fuel cell systems in FCEVs are assumed to be mass
produced under the 2DShighH2.
By 2050, power-to-power systems can achieve
mitigation costs of well below USD150 per tonne of
CO2. Depending on the regional context, they can
therefore be an attractive mitigation option under
the ETP2DS, since the CO2 price ranges between
USD150 and USD170 per tonne by that time.
From a marginal abatement cost perspective,
power-to-gas systems are the least promising
option. They have lower LCOE compared to power-
to-power systems, but since the blend share of

carbon-free hydrogen or synthetic methane (in case
of methanation) in the natural gas is only 5%, their
emission benefit compared to burning pure natural
gas is also limited. Since both blended natural
gas and pure natural gas are burned using similar
gas turbines, the abatement costs are simply the
ratio between the price difference and the carbon
intensity difference of blended natural gas (with
pure hydrogen or with synthetic methane) and pure
natural gas.
This again underlines the idea of power-to-gas
options being a transition technology, making use
of existing infrastructure.
Figure 27: M
arginal abatement costs of different hydrogen-based
VRE power integration applications in 2030 and 2050
Marginal abatement costs
USD/tCO2
1 200
1 000
2030
800
2050
600
400
2030 2DS carbon price
200
2050 2DS carbon price
0
Power-topower
Power-to-gas
Power-to-gas
blending
methanation
Power-tofuel
KEY POINT: Power-to-fuel applications offer the lowest marginal abatement costs in the long term.
52
Industry
Industrial direct CO2 emissions peak in 2020 under
the ETP2DS (IEA, 2015), and innovative low-carbon
processes become critical to achieving the 2DS in
the long term.
For DRI processes, the hydrogen-containing reducing

gases could be based on hydrogen with a lowcarbon footprint, if it was available at competitive
costs. A further 60MtCO2 emissions per year could
be mitigated if the estimated 132GNm of hydrogen
required for DRI processes in 2050 in the 2DS were
decarbonised (a minimum carbon content in the
reducing gas is needed to produce crude steel).
The steel industry offers great emissions mitigation

potential through improving energy efficiency,
phasing out outdated technologies, switching
existing processes to a lower-carbon fuel (e.g.
shifting coal to gas-based DRI), recycling more
steel and deploying innovative processes. These
measures lead to a reduction of almost 2Gt of
CO2 emissions per year by 2050 in the ETP 2DS
compared to the ETP 6DS (baseline scenario).
If, by 2025, commercial-scale demonstration was

successfully achieved, and all blast furnaces were
equipped with top gas recovery systems to recover
and recycle the hydrogen-containing blast furnace
gas after CO2 separation, almost 370MtCO2 per
year could be saved (0.3tCO2 per tonne of pig iron).
This only takes account of the lower consumption of
coke in the blast furnace, and ignores the possible
benefits of integrating carbon capture.
This mitigation potential is to some extent based on

the more effective use of hydrogen-containing offgases using them for reduction purposes rather
than as simple fuels. In some cases, integrating
the use of these off-gases in the iron ore reduction
process coincides with oxygen-rich conditions,
which in turn facilitates the implementation of
carbon capture. Hence there is a double mitigation
effect related to the implementation of these
alternative processes, resulting from lower energy
requirements, as fossil-based reducing agents are
displaced, and from direct CO2 sequestration.
Refining industry
Steel industry
Under the ETP 2DS, a greater penetration of natural

gas-based DRI compared to the 6DS (11% more)
enables emission savings of around 95MtCO2 per
year by 2050. In the case of DRI, a reducing gas that
contains the hydrogen needed to reduce the iron
ore is typically produced on site, either from natural
gas via SMR or from coal gasification.
In the post-2030 time frame, the deployment
of innovative processes, which to some extent
use hydrogen-containing gases as reducing
agents, leads to 624MtCO2 of annual direct
emission reductions by 2050 within the ETP2DS.
These emission reductions also include those
resulting from CCS, and build on the assumption
that successful demonstration is achieved on a
commercial scale.
In addition to the above-mentioned measures
included within the ETP2DS, other mitigation
options related to hydrogen technologies are
feasible in the steel industry. These comprise the
direct use of low-carbon footprint hydrogen or
the wider deployment of processes to recycle
hydrogen-containing gases.
Even under ambitious climate scenarios such as

the ETP2DShighH2, high energy density liquid
fuels for transport will remain in demand. By 2050,
petroleum-based fuels could still account for as
much as 60% of total transport fuel demand on
a global scale, with the market share of all liquid
fuels, including biofuels, standing at around 80%.
These high shares are largely due to the need for
energy-dense liquid fuels in road freight, air and
shipping, and also reflect the fact that alternative
vehicles using hydrogen or electricity have much
higher efficiencies, thus reducing their share of total
transport energy use.
As all liquid fuels require hydrogen during the
production process, its decarbonisation can have a
significant carbon mitigation impact. Around 2% of
the energy content of the final petroleum product is
needed in the form of hydrogen during the refining
processes (disregarding the fact that hydrogen
demand depends on the ratio of gasoline production
to distillate, the crude oil quality and many other
parameters). Replacing fossil hydrogen with lowcarbon footprint hydrogen, e.g. from methane
reformation with CCS, could lead to CO2 emission
reductions of some 100MtCO2 per year by 2050.
Chemical industry
Hydrogen is used as a feedstock in the synthesis
of high-demand chemicals such as ammonia and
methanol. The fossil-based steam-reforming process
used for hydrogen generation is one of the largest
energy-consuming steps in the synthesis of these
chemical products. Although shifting hydrogen
production from fossil to renewable-based routes
53
increases the energy intensity of the process, it

still is an attractive carbon mitigation option. A
30% replacement of fossil-derived hydrogen by
renewable alternatives by 2050 could save emissions
of 30MtCO2 per year in the chemical industry sector
(IEA, 2013).
of the FCEV fleet, the need for future supplies of

hydrogen for transport can be taken into account
when adding new SMR capacity to existing or new
refineries. Investment in slightly over-dimensioned
SMR capacity might be attractive in return for
potential future revenues from the sale of the
additional hydrogen production as transport fuel.
Synergies between
energy sectors
As the increase in commercial demand for

hydrogen with a low-carbon footprint from existing
applications (such as refining or steel production)
and from emerging applications occurs in certain
clusters or regions, the low-carbon price premium
for hydrogen from CCS-equipped plants can be
shared. Furthermore, multiple sources of demand
might enable arbitrage and management of hydrogen
production capacity in the face of variable demand.
This can raise capacity factors and competitiveness
and can improve the business case for polygeneration plants, such as IGCC plants that balance
the production of low-carbon hydrogen for electricity
generation with the supply of hydrogen for transport,
industry and electricity storage infrastructure.
Most importantly, the use of hydrogen allows for

the cross-sectoral integration of low value, surplus
renewable electricity in energy demand sectors
such as transport and industry. This enables the
further decarbonisation of these sectors while
unlocking new sources of system flexibility in the
power sector at the same time.
During the very early phase of FCEV market
introduction, hydrogen demand will need to be
covered by existing generation capacity or by
by-product hydrogen from the chemical and steel
industry, if physical properties such as hydrogen
concentration, pressure and purity allow for
economically viable conditioning to make the
hydrogen quality acceptable for use in PEMFCs.
Although SMR capacity at existing refineries might
not provide the significant additional capacity
necessary to supply hydrogen during the upscaling
Large numbers of FCEVs in the transport sector

will have an impact on the cost of PEM fuel cell and
possibly also PEM electrolyser stacks. This might even
be necessary to achieve the reductions in the cost of
PEM electrolysers envisaged in the Vision, as the
market for PEM electrolysers might on its own be
insufficient to realise the required economies of scale.
Box 9: Electrolysers in the control power market

PEM electrolysers are very flexible with respect
to ramp-up and load range cold start to full
power is possible in less than 10 seconds and
the dynamic range almost covers the entire
scale from 0% to 100% load, with loads of up to
300% possible over short times. The easy start
and stop procedures, without the need to purge
inert gases or for preheating, further increase
operational flexibility and reduce idle power
consumption. Additionally, transient operation
is not linked to faster degradation.
This behaviour enables PEM electrolysers to
be operated in a dynamic way and to use
arbitrage effects on the electricity spot market.
This can significantly reduce the LCOE of
hydrogen generation. When using electricity
for electrolysis along the increasing cumulative
54
average spot price, together with the respective

annual utilisation rate, an optimal annual load
factor and threshold price can be determined,
up to which electricity should be bought on the
spot market in order to minimise the LCOE of
hydrogen generation (Figure 28).
In that illustrative example, buying electricity
at up to USD 72 per MWh, and therefore
allowing the electrolyser to be operated for
more than 5 000 hours of the year (i.e. 56%
annual load factor), would reduce the LCOE of
hydrogen generation by almost 30% compared
to 100% utilisation of the electrolyser at an
average annual spot market electricity price of
USD69per MWh.
If the electrolyser was able to participate in the

primary control power market, where providing
negative controlling capacity is remunerated,
hydrogen generation costs could be further
reduced, while providing ancillary services to

the power system. The cost of hydrogen could
be reduced even more if by-product oxygen
was sold.
Figure 28: Electricity price arbitrage and hydrogen generation costs

250
250
200
200
150
150
30% Gain
100
100
50
50
0
0
-50
876
1752
2628
3504
4380
5256
Hours of the year
6132
7008
7884
8760
Levelised costs of electricity

USD per MWh
Levelised costs of hydrogen

USD per MWh
Hydrogen
Dynamic LCOE
Steady LCOE at 100%

annual ulizaon
Electricity
Spot price
Cumulave average spot
price
Point of opmal
operaon
-50
KEY POINT: A more dynamic use of the electrolyser with optimised operation with respect to input
electricity costs and annual utilisation rate can significantly reduce LCOE of hydrogen generation.
55
56
2030
Today
USD/
kWh
USD/
kWh
Investment
cost
conversion
Investment
cost
storage
Fixed
O&M
Life time hours or

years
Fixed
O&M
Efficiency
USD/
kWh
Investment
cost
storage
5%
870
95000
75%
5%
1150
USD/kW
Investment
cost
conversion
74%
75000
5%
800
75000
82%
5%
2600
40000
73%
AlkaPEM
line
electroelectrolyser
lyser
Life time hours or

years
Efficiency
Unit
3%
440
30
82%
3%
550
30
77%
NG
SMR
5%
700
30
73%
5%
1370
30
70%
NG SMR
with
CCS
5%
1280
30
57%
5%
1670
30
56%
Coal
CCS
5%
1320
30
50%
5%
4930
30
50%
Biomass
gasification
Hydrogengenerationandconversion
5%
450
20000
50%
5%
700
7000
50%
Alkaline FC
5%
830
80000
54%
5%
3200
60000
43%
PEM FC
5%
1620
75000
42%
5%
5800
40000
29%
H2 PtP
PEM/
PEM
5%
1700
75000
38%
5%
4350
60000
29%
H2 PtP
ALK/
PEM
5%
1420
75000
35%
5%
3230
40000
26%
H2 PtP
PEM/
OCGT
5%
1050
75000
82%
5%
2850
40000
73%
H2 PtG
PEM
HENG
5%
2280
75000
67%
5%
4090
40000
58%
H2 PtG
PEM
methan.
3%
50
1500
50
80%
3%
50
1500
50
80%
PHS
EnergystorageandVREintegration
5%
15
800
30
75%
5%
30
1000
30
60%
CAES
3%
500
30
45%
3%
500
30
39%
OCGT
Benchmark
Table 15: P
arameters used in the model for stationary hydrogen generation and conversion technologies
as well as for energy storage and VRE integration systems today and in the future
Parameters of key technologies today

and in the future as used in the model
57
USD/
kWh
USD/
kWh
Investment
cost
conversion
Investment
cost
storage
Fixed
O&M
Life time hours or

years
Efficiency
5%
700
95000
78%
5%
640
75000
86%
AlkaPEM
line
electroelectrolyser
lyser
3%
420
30
86%
NG
SMR
5%
670
30
77%
NG SMR
with
CCS
5%
1220
30
60%
Coal
CCS
5%
1250
30
53%
Biomass
gasification
5%
360
20000
53%
Alkaline FC
5%
660
80000
57%
PEM FC
5%
1300
75000
44%
H2 PtP
PEM/
PEM
5%
1360
75000
40%
H2 PtP
ALK/
PEM
5%
1140
75000
37%
H2 PtP
PEM/
OCGT
5%
840
75000
86%
H2 PtG
PEM
HENG
5%
1820
75000
71%
H2 PtG
PEM
methan.
3%
50
1500
50
80%
PHS
EnergystorageandVREintegration
5%
15
760
30
79%
CAES
3%
500
30
47%
OCGT
Benchmark
Sources: IEA data; Bnger, et al. (2014), Power-to-Gas (PtG) in Transport - Status Quo and Perspectives for Development; Decourt et al. (2014), Hydrogen-Based Energy Conversion,
More Than Storage: System Flexibility; Dodds and McDowall (2012), A Review of Hydrogen Production Technologies for Energy System Models; ETSAP (2014), Hydrogen Production and
Distribution; FCH-JU (2014), Development of Water Electrolysis in the European Union, Fuel Cells and Hydrogen Joint Undertaking; Giner Inc. (2013), PEM electrolyser incorporating an
advanced low-cost membrane, 2013 Hydrogen Program Annual Merit Review Meeting; Hydrogen Implementing Agreement Task 25 (2009), Alkaline Electrolysis; NETL (2013), Carbon
Dioxide Transport and Storage Costs in NETL Studies; NETL (2010), Production of High Purity Hydrogen from Domestic Coal: Assessing the Techno-Economic Impact of Emerging Technologies;
NREL (2010), NREL (2009a), Scenario development and analysis of hydrogen as a large-scale energy storage medium, RMEL Meeting; NREL (2009b), Scenario Development and Analysis
of Hydrogen as a Large-Scale Energy Storage Medium; Saur (2008), Wind-To-Hydrogen Project: Electrolyzer Capital Cost Study; Stolzenburg et al. (2014), Integration von Wind-WasserstoffSystemen in das Energiesystem - Abschlussbericht; Schaber, Steinke and Hamacher, (2013) Managing temporary oversupply from renewables efficiently: electricity storage versus energy
sector coupling in Germany, International Energy Workshop, Paris; US DOE (2014b), Hydrogen and Fuel Cells Program Record.
Notes: O&M = operation and maintenance; PtG = power-to-gas; PtP = power-to-power.
2050
Unit
Hydrogengenerationandconversion
Hydrogen technology development:

Actions and milestones
Technologies using low-carbon footprint hydrogen
can be valuable in various end-use applications,
notably in transport, but also for VRE integration.
They are particularly beneficial if low-carbon energy
needs to be stored, either in large quantities or under
space and weight restrictions in mobile applications.
Synergies between hydrogen end-use demand and
VRE integration can unlock the carbon emission
mitigation potential of otherwise-curtailed lowcarbon electricity. In combination with CCS, lowcarbon footprint hydrogen can also enable the use of
low-cost fossil resources in the transport sector.
The following section provides actions and

milestones together with indicative timelines, which
have been developed with experts to foster the
deployment of hydrogen technologies in the future.
Overarching actions and milestones have been
identified, which are followed by technology- or
application-specific development pathways, based
on metrics such as cost and efficiency targets,
which have been used to define the Vision.
Data assessment
and model development
This roadmap recommends the following actions
Development of appropriate modelling tools
Develop modelling tools that allow the investigation of problems

at the nexus between long-term and short-term optimisation of the
energy system. Incorporate high granularity with respect to time and
area in inter-sectoral energy system optimisation modelling environments. Investigate the impact of low utilisation rates in hydrogen
refuelling infrastructure on hydrogen costs in the transport sector.
2015-18
Address data gaps
Build a comprehensive and consistent dataset with high spatial

resolution, including regional resource endowments, renewable
energy potential, existing and planned energy T&D networks, and
geological formations suitable for hydrogen and carbon storage.
2015-18
Support R&D for system Support R&D projects that increase the understanding of the interintegration projects
actions between different energy sectors, and which help to quantify benefits and challenges of system integration beyond energy
flows, including questions relating to information flows, system
controllability and robustness, as well as data security aspects.
2015-18
Support coordination
between relevant actors
2015-18
Support initiatives to bring together relevant actors from different

parts of the energy system to coordinate technology development
and market introduction scenarios.
More detailed modelling tools are needed at a

national and international level to quantify, in robust
terms, the value of system integration against a
background of mitigating climate change and the
need to provide access to secure and economic
energy. These energy system optimisation tools need
to be: 1) integrated, which means that all energy
sectors (supply, transport, building and industry)
are optimised at the same time; 2) granular with
respect to temporal and spatial resolution; and
3)capable of capturing long-term and short-term
decision-making problems. Sectoral integration
is necessary to correctly represent the costs and
benefits of possible mitigation measures among the
58
Time frame
whole energy system, and to identify optimal energy

pathways across energy sectors. Temporal and spatial
granularity is needed to provide realistic results for
energy demand and supply, as well as to achieve
a more robust quantification of the need to shift
energy over time and space. For the latter, it is crucial
to improve representation of energy T&D networks
within modelling frameworks. A broader inclusion
of short-term operational aspects in energy system
optimisation is needed to better anticipate the use
of all flexibility options within the energy system,
according to their particular costs and benefits.
Technology development
Electrolysers
Time frame
Electrolysers in general
Optimise the technology with a focus on cost reduction. Key

areas of development comprise increased operational flexibility
by improving ramp-up rates, start times and stand-by energy use.
Draw upon the modularity of electrolysers and the respective
flexibility of power capacity.
Complete by
2020-30
PEM electrolysers
Reduce cost to USD 800 per kW through optimised manufacturing,

more resistant polymer membranes and reduced noble metal
content. Increase efficiency to more than 80% (HHV). Increase
lifetime to at least 50 000 hours. Increase stack capacity to multiple
MW. Increase total system capacity to the 100 MW scale. Achieve
ramp-up rates to comply with the primary control power market.
Complete by
2025-30
Alkaline electrolysers
Reduce investment cost to below USD 900 per kW. Increase

efficiency to more than 75% (HHV). Increase current density
through higher operating temperature and pressure. Reduce O&M
costs. Increase operational flexibility through reduction in minimal
load. Increase operating pressure to minimise subsequent need to
pressurise the hydrogen gas.
Complete by
2025-30
SO electrolysers
Prove commercial scale. Increase lifetime to at least 20 000 hours

at degradation rates below 8% per year. Achieve a minimum
operational flexibility to respond to future power market
requirements.
Complete by
2025-30
Electrolysers could be a pivotal technology for

achieving a wider deployment of low-carbon
footprint hydrogen in the energy system. They
can help to establish new links between the power
sector and transport, buildings and industry by
enabling new interconnections between the power
grid, the natural gas network and the transport
fuel infrastructure. Electrolysers can unlock the
potential of hydrogen technologies to contribute
to overall energy system flexibility. The generation
of renewable hydrogen using electrolysers will
be dependent on low-cost, renewable electricity

with constrained annual availability. It is therefore
particularly important to focus on reducing
investment cost to allow for capital recovery under
restricted annual full-load hours. Viable future
business models need to be based on benefit
stacking: the value of all possible by-products needs
to be realised. One option is to use electrolysers as a
primary control power to provide ancillary services
beyond the sole generation of hydrogen.
59
Fuel cells
Time frame
Fuel cells in general
Optimise both capital costs and efficiency. Efficiency is a key

parameter for stationary applications.
Complete
by 2025
PEMFCs, mobile
applications
Reduce real-world manufacturing costs to below USD 80 per kW

through optimised manufacturing and reduced need for precious
metal, while keeping lifetime to at least 5 000 hours. Reduce
sensitivity to hydrogen impurities.
Complete
by 2025
PEMFCs, stationary
applications
Reduce investment cost to below USD 800 per kW by reducing

both the cost of the stack and the cost of balance of plant. Increase
system efficiencies to at least 50%. Increase lifetime to above
80000 hours. Reduce sensitivity to hydrogen impurities and prove
feasibility at large stack capacities. Achieve megawatt scale.
Complete
by 2025-30
Alkaline fuel cells
Increase technical lifetime to more than 10 000 hours.
Complete
by 2025-30
SOFCs
Increase cell lifetimes under real world conditions at acceptable

degradation to more than 50000hours. Improve operational
flexibility. Reduce investment costs to below USD 2 000 per kW.
Complete
by 2025-35
Fuel cells are a key technology to efficiently use

hydrogen as a high value energy carrier, especially
where small and medium-sized power outputs are
required. Unlike for electrolysers, the efficiency of
low-temperature fuel cells remains rather low and
needs to be improved. A better understanding of
dynamic degradation processes of the catalyst in
PEMFCs is needed to achieve high efficiencies over
the entire lifetime. While high-temperature fuel cells
already show promising efficiencies, current lifetime
and cost levels impede commercial application.
The focus of technology development depends on
the application: the focus for stationary fuel cells
is increasing efficiency and ensuring utilisation of

waste heat; the focus for mobile applications is on
cost reduction and increased lifetime.
Large-scale manufacturing processes are needed to
unlock the potential for cost reductions. In PEMFCs,
the membrane electrode assembly (MEA) accounts
for around 30% of the total system cost (Decourt
et al., 2014). With automated manufacturing, MEA
costs are reported to decrease significantly. It needs
to be proven in practice whether cost reductions
in PEMFC manufacturing can spill over into PEM
electrolyser production.
Hydrogen storage
Time frame
Underground hydrogen
storage
Establish national inventories of underground caverns suitable for

hydrogen storage. Develop demonstration projects for hydrogen
storage in salt caverns and prove feasibility to reduce investment
costs for storage to USD 1 per kWh. Prove the feasibility of
hydrogen storage in depleted oil and gas fields as well as aquifers.
Complete
by 2025-35
Pressurised tanks
Reduce material costs for high-pressure tanks on board FCEVs to at

least USD 15 per kWh.
Complete by
2025
Cryogenic storage and

Improve the efficiency of the liquefaction process to reduce energy
liquefaction of hydrogen losses to below 30%. Reduce boil-off through improved insulation
of the vessel as well as increased pressure levels.
Metal hydrides and
carbon nano-structures
60
Ensure continued R&D funding to further explore the potential

application of solid hydrogen storage options.
Complete
by 2030-35
2015
onwards
While medium-sized hydrogen storage using

pressurised steel vessels is a mature technology,
other small- and large-scale hydrogen storage
options still need further development. In the
case of on-board hydrogen storage for FCEVs,
the 70MPa high-pressure tanks are likely to be a
major cost factor in the future. Unlike for the fuel
cell stack, the costs of the tank are dominated by
material costs rather than manufacturing costs.
In the case of large-scale hydrogen storage, in
the near term the focus needs to be on improving
understanding of the geographically available
storage potential. Although salt caverns might
be superior from a technological point of view,
alternatives using depleted oil and gas reservoirs,
as well as aquifer formations, need to be further
investigated as options for storage potential. These

alternatives might be more useful due to their
geographical distribution and compatibility with
existing infrastructure.
Options for combining applications need to be
investigated and demonstrated in the near term to
facilitate large-scale, long-term energy storage, in
order to increase annual full-load hours of system
components. So-called bi- or even tri-generation
applications, producing hydrogen for transport in
addition to generating power and, in the case of trigeneration, also heat, can be the key to achieving
necessary asset utilisation. Participation in different
energy markets might be a prerequisite to making
hydrogen-based energy storage technologies
economically feasible.
FCEVs
Time frame
Investment costs
Achieve a price premium of 15% or less compared to hybridised

ICE vehicles at higher volume annual production rates.
Complete
by 2025
On-board hydrogen
storage
Reduce the volume and the weight of the hydrogen tank. Reduce
specific costs to at least below USD 15 per kWh.
Complete
by 2025
Fuel economy
and range
Achieve an on-road fuel efficiency of 0.8 kg of hydrogen per

100km.
Complete
by 2025
Improve on-road fuel efficiency to up 0.6 kg of hydrogen per

100km to reduce the size of the tank while achieving at least
500km range.
Complete
by 2035
Establish an international standard for refuelling pressure and shape

of the nozzle.
Complete
by 2020
Refuelling
The next ten years will be crucial for demonstrating

the large-scale mobility potential of FCEVs.
Although some manufacturers announced the
introduction of commercially available FCEVs
during 2015, it will be necessary to sell the first
tens of thousands of FCEVs to technophile first
movers around the globe to learn about consumer
acceptance and technology behaviour under reallife conditions.
than 15% above that for conventional hybrid cars,

taking into account the higher cost of the fuel. To
reduce the cost of the vehicle as a whole, each
subsystem needs to contribute, with the actual fuel
cell system being only one part of that. Being able
to equip otherwise-conventional vehicles with a
fuel cell powertrain and hydrogen storage will be a
major step towards larger-scale commercialisation.
Vehicle cost reductions are crucial to achieving

such an ambitious target. For success beyond the
large-scale technology demonstration phase, the
purchase price of FCEVs should not be much higher
61
Hydrogen T&D
and retail infrastructure
Tractor-trailer
combinations
Increase the capacity of tube trailers for transport of gaseous

hydrogen to above 900 kg. Increase the pressure to reduce the
need for compression work at the station.
Retail station
Define optimal hydrogen station layout with respect to hydrogen

phase (gaseous vs. liquefied), size, pressure and compression
scheme, taking into account region-specific characteristics of
hydrogen generation pathways. Define standardised refuelling
pressures. Consider proposals for modular or mobile hydrogen
stations to reduce under-utilisation during FCEV market
introduction and scale-up. Reduce station area footprint. Design
user-friendly and standardised dispensers. Reduce investment
costs to below USD 1 million for small stations dispensing in the
region of 200 kg of hydrogen per day.
Complete
by 2020-25
Compressor
Eliminate uncertainties and focus on decreasing costs for

compression. Achieve investment costs of USD 300 per kW
of hydrogen throughput and less for an 88 MPa compressor.
Develop scenarios to determine optimal compression levels
throughout each stage, from hydrogen generation to retail at the
station.
Complete
by 2020-25
The build-up to a minimum hydrogen T&D and

retail network will be the main barrier to the
widespread use of FCEVs in transport. Large-scale
demonstration programmes, initially bringing
several thousand FCEVs on the road, must be
supported by hydrogen refuelling networks
providing coverage in selected core regions.
Furthermore, these demonstration regions need to
be connected via corridors to enable first movers
to use their FCEV on long-distance trips, and
should be developed based on existing hydrogen
infrastructure. Various clusters are currently planned
in California, Germany, the UnitedKingdom, France,
theNetherlands, Japan and Korea.
Certain key parameters, such as vehicle on-board
storage pressure and the shape of the refuelling
nozzle, need to converge in the next decade to
facilitate infrastructure development. Furthermore,
station layouts that allow for modular expansion of
refuelling capacity alongside demand need to be
62
Time frame
Complete
by 2025
developed, in order to minimise under-utilisation.

The set-up of this entirely new energy infrastructure
will not be based on a single approach, but
all options, including vehicle fleets and public
transport, need to be integrated to create sufficient
hydrogen demand around the initial T&D clusters.
Finally, scaling up from clustered hydrogen retailing
to national and regional coverage will demand
major investment supported by government
programmes, and will require consensus among
a great number of stakeholders, from the oil and
gas industry, utilities and power grid providers, car
manufacturers, and local, regional and national
authorities. Achieving this common understanding
of future development might well be the most
serious hurdle to overcome.
CO2 capture and storage

Time frame
CO2 capture from SMRs
Raise the number of operating SMRs equipped with large-scale

CO2 capture (e.g. 100 000 tonnes of CO2 per year [tCO2/yr] and
above) to five worldwide. Public funding should support the use of
capture technologies that promise lower costs and higher capture
rates, such as cryogenics, vacuum pressure swing adsorption
(VPSA) and membranes.
Complete
by 2020-25
Poly-generation
with CCS
Demonstration of commercial poly-generation of low-carbon

hydrogen and other commodities (electricity, urea, methanol) from
coal conversion combined with CCS in five large-scale projects
worldwide.
Complete
by 2020-25
CO2 capture and supply
Reduce the cost of CO2 capture from flue gases and other sources
identified as potential CO2 suppliers for power-to-gas and powerto-liquids processes to USD 15 to USD 50 per tonne of CO2.
Suppliers could include biogas upgraders, bioethanol mills, steel
plants, refineries, chemical plants, power plants or direct air
capture (depending on timing of anticipated need).
Complete
by 2025-35
CO2 storage
Implement policies that encourage storage exploration and

characterisation, and development of CO2 storage resources in
countries where hydrogen production from fossil fuels with CCS is a
cost-effective option. To manage multiple emission sources, public
and private investment in strategic CO2 storage assets needs to be
increased from todays low levels in most countries, in parallel with
stimulating the emergence of viable CO2 storage service providers
and establishing governance frameworks that ensure safe and
effective storage.
Complete
by 2020-30
While CO2 capture from SMR currently operates at

scales of 1million tCO2/yr per plant, or 1.2billion
standard cubic feet per day (BSCFD) of hydrogen,
further action is needed to make the full CCS
chain a financeable proposition for climate change
mitigation. The cost of CO2 capture from SMR
can be further reduced. The sale of low-carbon
hydrogen from coal, alongside products destined
for other commodity markets using polygeneration, needs to be demonstrated. Above all,
however, incentivising permanent CO2 storage is
the key to unlocking the cost advantage presented
by hydrogen production coupled with CCS. The

IEA Technology Roadmap for Carbon Capture and
Storage (2013) recommended fivekey actions for
CO2 storage in the near term. These relate primarily
to policies and regulations that encourage CO2
storage resources to be characterised and made
commercially available. Ensuring safe and effective
storage, sound management of natural resources,
and public consultation in line with best practice are
also essential items.
63
Policy, regulatory framework and finance:

Actions and milestones
Strong policies are needed if hydrogen as an energy
carrier is to play a major role in a future low-carbon
energy system. The large-scale deployment of
hydrogen is linked to the introduction of entirely
new technologies, both on the energy supply side
and the demand side, requiring the establishment
of a new energy T&D and retail system in parallel.
The simultaneous development of such complex
tasks will require proactive intervention and
coordination.
This roadmap highlights several specific challenges
for policymakers. These include the reduction of
emissions from road transport, the facilitation of
high levels of variable renewable electricity and
the creation of markets for low-carbon industrial
production based on increased use of hydrogen
and fuel cell technologies. Sound policy approaches
will be needed to stimulate effective competition
between technological solutions, including but
not limited to the use of hydrogen and fuel cells.
In addition to internalising the environmental and
social costs of GHG emissions, policy can provide
directed support to promising technologies to
reduce costs, improve performance and enable
early market introduction. Preferably, such measures
should be time-limited and encourage low-carbon
options to compete on their merits.
Governments can act as catalysts to speed up
developments by providing support in the form
of RD&D funding, access to attractive financing
programmes, and the necessary regulatory and

policy framework. The latter point is especially
important as governments need to take the lead on
providing a stable investment environment, clearly
formulating long-term targets, especially with
respect to energy use and climate change.
Mobilising private capital is a prerequisite for
the large scale deployment of hydrogen and fuel
cell technology. During the past decade, several
initiatives and public-private partnerships have
been created to coordinate action between
stakeholders and to secure funding (Table 16). For
example, to develop the hydrogen generation and
refuelling infrastructure necessary for the successful
introduction of FCEVs, car manufacturers, fuel cell
and electrolyser producers, oil, gas and power
suppliers, as well as transport service providers,
have created common initiatives to try to manage
the investment risk. Ultimately, the success of
these initiatives will be measured by their ability to
achieve binding agreements among the different
stakeholders to tackle the chicken and egg
problem. Globally, significant annual funding, in the
order of several hundred million USdollars, is being
spent on hydrogen and fuel cell technologies as well
as related infrastructure development13.
13. F
or comparison, in 2012 about USD1billion has been
spent by governments on solar and CSP RD&D, and around
USD1.5billion was allocated to biofuels.
Table 16: I nitiatives and public-private partnerships

to promote hydrogen and fuel cell technologies
64
Region
Exemple
Europe
zz
zz
zz
zz
zz
zz
zz
zz
North
America
zz CaFCP,California)
zz H2USA (UnitedStates)
zz Canadian Hydrogen and Fuel Cell Association (CHFCA, Canada)
Japan
zz The Research Association of Hydrogen Supply/Utilization Technology (HySUT)

zz Fuel Cell Commercialisation Conference of Japan (FCCJ)
Fuel Cell and Hydrogen Joint Undertaking (FCH-JU, EU)

Nationale Organisation Wasserstoff- und Brennstoffzellentechnologie (NOW GmbH, Germany)
Clean Energy Partnership (CEP, Germany)
Mobilit Hydrogne (France)
Scandinavian H2 Highway Partnership (SHHP, Scandinavia)
HyNor (Norway)
Hydrogen Sweden (formerly HyFuture, Sweden)
UK H2 Mobility (UnitedKingdom)
Policy
Target group
Technology
neutral
CO2-based vehicle taxation
Consumers
Feebate schemes
Consumers
Labelling schemes
Consumers
Vehicle perks free use of public parking, use of high

occupancy vehicle lanes, use of bus lanes, exemption
from road tolls
Consumer
Fuel economy standards
Car manufacturers
Zero-emission vehicle regulation
Car manufacturers
Low-carbon/renewable fuel regulation
Fuel suppliers
Direct vehicle purchase subsidies
Consumers
Vehicle purchase and fuel tax exemption
Consumers
Subsidies for H2 infrastructure
Policy measures to support the large-scale

application of FCEVs in transport can be categorised
by the target group they address, e.g. the consumer,
the car manufacturers or the fuel suppliers. They can
furthermore be distinguished as technology-specific
and technology-neutral support instruments.
A whole range of technology-neutral policies
exist that apply to the consumer and which can
be beneficial to the introduction of FCEVs. These
policies comprise annual vehicle taxation schemes
(e.g. in Germany) based on vehicle CO2 emissions,
or the introduction of feebate schemes for vehicle
registration taxes (e.g. in France). As hydrogen
vehicles would certainly fall within the category
of low-emission vehicles, they would benefit from
lower taxation or higher rebates. Together with
labelling schemes clearly stating fuel economy and
CO2 emissions (based on region-specific emission
factors for hydrogen and electricity) and other soft
measures to incentivise low-emission vehicles, such
as free use of public parking spaces, the use of bus
lanes and high occupancy vehicle lanes or the free
use of toll roads, these technology-neutral measures
can already contribute to attracting consumer
interest in FCEVs.
Technology-neutral policies such as demanding,
long-term fuel economy standards can be a
strong incentive to car manufacturers to introduce
Time frame
Implement
by 201520
Fuel suppliers
low-emission vehicles, as they can significantly

contribute to achieving corporate average
fuel economy targets. The same is true for the
introduction of zero-emission vehicle quotas for
government fleets, as formulated for example in the
Californian ZEV Action Plan (Governors Interagency
Working Group on Zero-emission Vehicles, 2013).
On the fuel supply side, incentivising the
introduction of low-carbon fuels can be broadened
to hydrogen. If, for example, hydrogen were to
qualify as a biofuel under the US Renewable Fuel
Standard 2 (RFS 2), a strong incentive to scale up
hydrogen generation capacity would be provided.
Similarly, the EU Fuel Quality Directive could
incentivise the deployment of low-carbon hydrogen
throughout the production process of conventional
petroleum fuels.
Due to the complexity of the value chain,
coordinated policy support may be needed
simultaneously in a number of areas. Alongside
technology-specific support for FCEVs, the scale
up of hydrogen generation, T&D and retail
infrastructure will be necessary for wider market
adoption.
In Europe, the TEN-T programme was established to
support the construction and upgrade of transport
infrastructure across the European Union. Part of
65
this program is the HIT (Hydrogen Infrastructure

for Transport) project, aiming at establishing a
basic network of European hydrogen refuelling
infrastructure to enable large distance travel using
FCEVs. The recently adopted Directive on the
deployment of alternative fuels infrastructure
acknowledges the need for the built-up of hydrogen
refuelling infrastructure as a prerequisite for FCEV
deployment. It furthermore concludes that by
end 2025 an appropriate number of hydrogen
refuelling stations needs to be in place within those
Members States which adopted the use of hydrogen
for road transport as one of their national polices.
In addition to this, FCEVs and hydrogen will need to
be exempted from taxation of vehicle registration,
vehicle ownership and purchase of fuel, to close
the gap in TCD with conventional cars. Depending
on the pace of FCEV market uptake, these tax
exemptions might be necessary for a time period of
at least 10 to 15years after FCEV market introduction,
and should be closely monitored and regularly
adjusted to prevent over- or underspending.
Additionally, direct subsidies to reduce the

remaining gap in TCD tax may be necessary for
an extended time period, if cost parity with
conventional technology is envisaged. These direct
subsidies need to be split among consumers, car
manufacturers and fuel suppliers in a way that
the market is stimulated and the investment risk
for both car manufacturers and fuel suppliers is
minimised. This might involve direct subsidies to
consumers when buying an FCEV, government
support to car manufacturers to scale up FCEV
manufacturing, and government support
to the fuel suppliers to help set up an initial
refuelling infrastructure. All kinds of subsidies
must be thoroughly monitored and adapted to
market conditions in order to prevent over- or
underspending. As highlighted in the Vision, TCD
can be an effective measure to evaluate and adjust
levels for direct subsidies in transport on a regular
basis. Particularly for technologies expected to have
high learning rates, determining the right amount of
subsidies can save resources.
Hydrogen in
stationary applications
Policy
Target group
Long-term emission reduction targets
Carbon pricing
Incentivise VRE operators to adopt grid

integration measures
Utilities, decentralised generation
Increase price transparency for power

generation and heat production
Utilities, decentralised generation,

energy storage operators
Facilitate entry into energy markets
Decentralised generation, energy

storage operators
Enable benefit-stacking for energy storage

systems.
Energy storage operators
Exemption of electrolysers from

renewable surcharge and grid usage fees
Utilities, energy storage operators,

grid operators
Green gas certificates
Time frame
Implement
by 201520
Energy storage operators
Similar to mobile applications, policies to incentivise

hydrogen in stationary applications can be divided
into technology-neutral and technology-specific
measures.
66
Technology
neutral
As for any other low-carbon technology, the

presence of long-term emission reduction targets is
a prerequisite for hydrogen technology deployment.
In combination with an increasing carbon price as
a result of tightening emission targets, stationary

hydrogen-based energy technologies can become
competitive in the future if finance for RD&D is
secured during the early phase of technology
development.
This basically means that hydrogen-based energy

storage systems should be able to be remunerated
for all ancillary services they provide to the grid.
The further development of green gas certificates
can provide the option of selling low-carbon gas
to consumers who are prepared to pay a price
premium. This also opens up the possibility for PtG
operations to sell lower carbon footprint gas to
consumers, who are not physically connected to the
same natural gas distribution grid.
It is necessary to incentivise not only the addition of

VRE capacity, but also its integration into the power
system to prevent increasing rates of curtailment
and to reflect the real costs of VRE. This will foster
the uptake of flexibility measures within the energy
system, and thus possibly also the deployment of
hydrogen-based energy storage systems.
In 2014, a new proposal for a Council Directive on

calculation methods and reporting requirements
pursuant the European Fuel Quality Directive
(European Commission, 2014) was presented. It
lays out a method for fuel suppliers to calculate
the lifecycle GHG emissions associated with their
products, in order to reach the 6% emission
reduction target, which is due to be achieved by
2020. This, together with post-2020 plans to achieve
a 60% emission reduction in transport within the
EuropeanUnion by 2050 (European Commission,
2011), can finally incentivise the use of low-carbon
hydrogen in the refinery sector.
Hydrogen technology-specific policy measures

could facilitate the technologys entry into energy
markets, such as the exemption of electrolysers
from renewable surcharges and grid usage fees.
This is justified given the fact that the potential use
of electrolysers in the primary control power market
would actually help to integrate otherwise-curtailed
electricity from VRE, and therefore ease pressure
on power grids. It will furthermore be essential to
allow energy markets to qualify for benefit stacking.
The role of codes

and standards
Time frame
Development of a methodology to include region-specific upstream emissions during fuel

production within the new Worldwide harmonised Light vehicles Test Procedures (WLTP).
Establish a performance-based global technical regulation for type approval of motor
vehicles within the UN framework to ensure safety of FCEVs being comparable or superior to
those of conventional PLDVs
Implement
by 2015-20
Hydrogen handling security regulation

Hydrogen metering regulation
Hydrogen refuelling equipment standardisation
Natural gas-hydrogen blend quality and safety regulation
Determination of maximum blend shares for hydrogen in natural gas by application
Standardisation is an important element on the way

towards large-scale application of hydrogen as an
energy carrier. In particular, the security regulations
for FCEVs and hydrogen handling in the transport
sector need to be harmonised on a global scale.
In 2014, the UN Working Party on Passive Safety
submitted a Proposal for a new Regulation on
hydrogen and fuel cell vehicles (HFCVs) (Working
Party on Passive Safety, 2014) to establish uniform

provisions for the type approval of FCEVs. This work
needs to be finalised to allow for the sale of large
volumes of FCEVs.
Currently, requirements for FCEVs are developed
with reference to the WLTP, which is being
developed to determine the levels of pollutants and
CO2 emitted by new PLDVs and light commercial
67
vehicles in a globally harmonised way. Setting up a

sound methodology to include emissions during the
fuel production process is a necessary step towards
measuring CO2 emissions from FCEVs, BEVs and
plug-in hybrids.
The harmonisation of standards for hydrogen
metering at refuelling stations can help reduce the
cost of developing station equipment. Establishing
global standards is a matter of urgency for other
hydrogen refuelling station equipment such as
the dispenser, including the nozzle to connect the
dispenser to the car.
To enable PtG applications, quality standards

for blended natural gas need to be developed to
enable safe operation of natural gas-fuelled end-use
applications, as well as to allow for correct metering
on an energy content basis. This is required for
blending hydrogen in local distribution grids and
for selling blended natural gas through transmission
lines on a regional, national or even international
scale. As first step, international agreement is needed
on maximum blend shares acceptable for different
end-use applications.
Finance
Time frame
Provision of long-term low-interest loans

Development of renewable energy grants and funds
Development of green bonds
Investment tax credits
Implement
by 2015-20
Long-term RD&D funding

University funding, competitive awards
Securing finance for innovative technologies is

often challenging. Both governments and financial
institutions are essential to providing access to
necessary funds and to incentivise investment in
low-carbon energy technologies.
the technology. They will need a stable, longterm support to deployment and the provision of
long-term, low-interest loans or the development
of renewable energy grants and funds can help to
reduce the costs of capital.
Government support needs to be adapted to the

different phases of the innovation and deployment
cycle and the right support depends on the
maturity of the technology and the degree of
market uptake (IEA, 2015).
New financing mechanisms such as green bonds

can also be a mean to lower the costs of capital.
The first green bonds14 were issued in 2008 by the
World Bank Treasury and by July 2014, green bond
issuances well exceeded USD 20 billion twice the
amount as those issued in 2013 (World Bank, 2014).
For technologies at the earlier stages of the

innovation cycle, such as high-temperature fuel
cells and electrolysers, which need to substantially
improve performance and reduce costs to achieve
technical and economic viability, technology push
mechanisms are most effective. Securing long-term
RD&D funding, e.g. through research grants, is a
prerequisite for successful upscaling of hydrogen
and fuel cell technologies. Apart from financing
research, the tendering of competitive awards can
be an attractive option.
Technology innovation for post-commercialisation
deployment is largely based on mobilising private
investments from the industries manufacturing
68
Furthermore, investment into hydrogen and fuel

cell technologies qualifying as low-carbon energy
technologies can also be incentivised through
special tax programmes, aimed at reducing tax
liabilities on corporate or income taxes of businesses
and households.
14.
Green bonds are fixed-income, liquid financial instruments
that are used to raise funds dedicated to climate mitigation,
adaptation and other environment-friendly projects (www.
worldbank.org/en/topic/climatechange/brief/green-bondsclimate-finance).
In general, all financial instruments need to have

transparent methodologies for deciding on the
qualification of energy technologies under these
programmes. In this respect, current and expected
abatement costs are a useful measure to compare
clean energy technologies, as they allow the costs of
carbon mitigation to be directly compared.
significant role beyond the regions discussed in this

roadmap. The results of costly learning processes
need to be accessible globally. For countries such
as China or India, the development of hydrogen
technologies in combination with CCS could be
attractive to transform abundant domestic fossil
resources into low-carbon transport fuels.
International collaboration
Social acceptance and safety
International collaboration is key to successful

technology development programmes. In
developed regions, replacing parallel development
of work streams with coordinated RD&D efforts can
contribute significantly to reducing timescales and
optimising resources. This is especially true in times
of tight public funding budgets. Platforms such as
the hydrogen-related Implementing Agreements
within the IEA Technology Network (e.g. Hydrogen
Implementing Agreement [HIA] and Advanced
Fuel Cell Implementing Agreement [AFCIA]) or the
International Partnership for Hydrogen and Fuel
Cells (IPHE) need to be used in an efficient manner
to deepen international teamwork.
Effective public education will be essential to

the widespread social acceptance of hydrogen
technologies. Convincing the consumer that FCEVs
are safe will be one of the major tasks during the
early market introduction phase. Early education
of all relevant stakeholders, including ambulance
and fire service personnel, is critical. This can be
done through continued information campaigns
and, in respect of safety-related matters, through
the further development of international hydrogen
technology-related training programmes, such as
the European HySafe project.
International cooperation can successfully

engage emerging economies in activities that can
enhance their domestic technological capability
(or absorptive capacity) to deploy clean energy
technologies and also to deliver clean energy
innovation autonomously. Knowledge spill-over
effects between developed and developing regions
are necessary if hydrogen technologies are to play a
Furthermore, the results of FCEV crash tests (e.g.

those crash tests required for the NCAP safety
rating) should be disseminated through information
campaigns. Adequate training of hydrogen
refuelling station personnel and reassuring
operation of the refuelling station equipment are
preconditions to reducing security concerns.
Policy, regulatory framework and finance: actions and milestones
69
Conclusion:
Near-term actions for stakeholders
This roadmap investigates the potential for
hydrogen technologies to help achieve an emission
trajectory needed to limit the long-term global
average temperature rise to 2C. It includes specific
milestones that the international community can
use to track the progress of hydrogen technology
deployment, if hydrogen is to play a significant
role as an energy carrier by 2050, as outlined in the
Lead
stakeholder
2DShighH2. The IEA, together with governments,

industry and other key stakeholders, will report
regularly on this progress, and recommend
adjustments to the roadmap as needed.
Recommended actions for key stakeholders are
summarised below, and are presented to indicate
who should take the lead in such efforts.

zz P
ush forward long-term climate targets and establish a stable policy and regulatory
framework, including for example carbon pricing, feed in tariffs or renewable fuel
standards to encourage fuel efficiency and low greenhouse gas emission technologies
across all energy sectors.
zz S
trengthen fuel economy and CO2 emission regulation as well as pollutant emission
standards for road vehicles beyond the time frames and modes already covered by
todays approaches.
zz A
pply monetary measures to incentivise alternative fuel vehicles, e.g. feebate systems,
CO2-based vehicle ownership and circulation taxation.
zz Improve and strengthen the policy framework to address upstream emissions during fuel
generation in the transport sector.
zz E stablish a power market framework, which allows for the adequate remuneration of all
power system services provided by energy storage technologies.
armonise safety codes and standards for hydrogen T&D and retail infrastructure as well
Governments zz H
as for hydrogen metering.
zz W
here regionally relevant, establish standards for natural gas quality with hydrogen
blend share.
zz S
upport research projects that increase understanding of the interactions between
different energy sectors, and which help to quantify benefits and challenges of system
integration.
zz S
upport RD&D necessary to improve key hydrogen conversion technologies such as
electrolysers and fuel cells.
zz S
upport government involvement in demonstration projects, especially with respect to
hydrogen transmission, distribution and retail infrastructure roll-out.
zz A
ddress potential market barriers where opportunities exist for the use of low-carbon
hydrogen in industry (e.g. in refineries).
zz E xtend information campaigns and educational programs to increase awareness-raising.
zz I dentify the lowest-cost system design and manufacturing methods for fuel cells and
electrolysers. Optimise lifetime and degradation and scale up system size.
zz D
emonstrate the large-scale mobility potential of FCEVs by proving on-road practicality
and economics across the supply chain of FCEVs. Put the first tens of thousands of FCEVs
on the road.
Industry
zz P
rove the economic feasibility and built-up hydrogen generation, T&D and retail capacity
necessary to refuel several tens of thousands of FCEVs.
zz Demonstrate hydrogen-based energy storage systems in large-scale applications.
zz W
here regionally relevant, accelerate activities directed at developing the capture and
storage of CO2 from fossil-derived hydrogen production into mature business activities.
zz Bring down costs and of FC micro combined heat and power systems.
70
Lead
stakeholder

zz P
rovide the tools to analyse the energy system, including all energy demand and
energy supply sectors, with the temporal and spatial resolution necessary to adequately
examine synergies between hydrogen demand, VRE integration and energy storage.
zz Improve the data on resource availability, costs and geologic formations suitable for
underground storage of gaseous energy carriers.
Academia
zz D
evelop strategies to cluster hydrogen refuelling infrastructure during technology
roll-out.
zz Include and improve linkages between different energy infrastructure systems (e.g.the
power grid and the natural gas grid) in national energy system models.
zz Improve methods to quantify directly and indirectly occurring upstream GHG emissions
during transport fuel generation, T&D and retail beyond the focus on carbon dioxide
emissions.
zz D
etermine maximum acceptable blend shares of hydrogen in natural gas to comply with
different end-use specifications.
Conclusion: near-term actions for stakeholders
71
Abbreviations, acronyms
and units of measurement
Abbreviations and acronyms
SOFC
solid oxide fuel cell
SR
smelt reduction
T&D
transmission and distribution
TCD
total costs of driving
ALK
alkaline
BEV
battery electric vehicle
BF
blast furnace
BFG
blast furnace gas
BOFG
basic oxygen furnace gas
BOP
balance of plant
CAES
compressed air energy storage
CCGT
combined cycle gas turbine
CCS
carbon capture and storage
EJ
Exajoule
CNG
compressed natural gas
Gt
Gigatonne
COG
coke oven gas
Kg
Kilogramm
CV
commercial vehicle
Km
Kilometre
DRI
direct reduced iron
kW
Kilowatt
EAF
electric arc furnace
Lge
Litre of gasoline equivalent
EL
electrolyser
MPa
Megapascal
ETP
Energy Technology Perspectives
Mt
Megatonne
FC
fuel cell
MW
Megawatt
FCEV
fuel cell electric vehicle
MWh Megawatt hour
HENG hydrogen-enriched natural gas

HFT
heavy freight truck
HHV
higher heating value
ICE
internal combustion engine
IGCC
integrated gasification combined cycle
LCOE
levelised cost of energy
ULCOS ultra-low-carbon dioxide steelmaking

VRE
variable renewable energy
WTW
well-to-wheel
Units of measure
TWh
Terawatt hour
LCOH2 levelised cost of hydrogen

LCV
light commercial vehicle
LHV
lower heating value
MCFC molten carbonate fuel cell

MEA
membrane electrode assembly
MFT
medium freight truck
NG
natural gas
OCGT open-cycle gas turbine

O&M operation and maintenance
PAFC
phosphoric acid fuel cell
PEM
proton exchange membrane
PEMFC proton exchange membrane fuel cell

PHEV
plug-in hybrid electric vehicle
PHS
pumped hydro energy storage
PLDV
passenger light-duty vehicle
PtG
power-to-gas
PtP
power-to-power
RD&D research development and demonstration

SMR
72
steam methane reforming
References
Argonne National LaboratoryNuclear Division
(2010), Technical Assessment of Compressed
Hydrogen Storage Tank Systems for Automotive
Applications, Argonne National Laboratory, Lemont,
IL.
Blum, L. et al. (2014), Overview on the Jlich SOFC
development status, 11th European SOFC and SOE
Forum, Lucerne.
Brmmer, T. (2013), Gas Quality and Interoperability,
Gasunie, Brussels.
Bnger, U. et al. (2014), Power-to-Gas (PtG)
in Transport - Status Quo and Perspectives for
Development, Deutsches Zentrum fr Luft- und
Raumfahrt e.V., Berlin.
Decourt, B. et al. (2014), Hydrogen-Based Energy
Conversion, More Than Storage: System Flexibility, SBC
Energy Institute, Paris.
Deutscher Verein des Gas- und Wasserfaches (2013),
Entwicklung von Modularen Konzepten zur Erzeugung,
Speicherung und Einspeisung von Wasserstoff und
Methan in Erdgasnetz, Deutscher Verein des Gasund Wasserfaches, Bonn.
Dodds, P. and W. McDowall (2012), A Review of
Hydrogen Production Technologies for Energy System
Models, UCL Energy Institute, University College
London.
Elgowainy, A. (2014), Hydrogen infrastructure
analysis in early markets of FCEVs, IEA Hydrogen
Roadmap North America Workshop, Argonne
National Laboratory, Bethesda, MD.
ETSAP (2014), Hydrogen Production and Distribution,
IEA Energy Technology Network, Energy Technology
System Analysis Programme.
Eurogas (2014), Statistical Report 2014, Eurogas,
Brussels.
European Commission (2014), Council Directive on
Laying Down Calculation Methods and Reporting
Requirements Pursuant to Directive 98/70/EC of the
European Parliament and of the Council Relating
to the Quality of Petrol and Diesel fuels, European
Commission, Brussels.
European Commission (2011), White Paper: Roadmap
to a Single European Transport Area Towards a
Competitive and Resource Efficient Transport System,
European Commission, Brussels.
European Commission (2000), Best Available

Techniques Reference Document on the Production of
Iron and Steel, Joint Research Centre, Brussels.
EVI (2015), Global EV Outlook 2015, Electric Vehicle
Initiative, Paris.
FCCJ (2015), Fuel Cell Commercialisation
Conference in Japan (FCCJ), http://fccj.jp/hystation/
index.html#hystop, (accessed 30 April 2015).
FCH-JU (2014), Development of Water Electrolysis in
the European Union, Fuel Cells and Hydrogen Joint
Undertaking.
Gahleitner, G. (2013), Hydrogen from renewable
electricity: an international review of power-to-gas
pilot plants for stationary applications, Hydrogen
Energy, pp. 2039-2061.
Giner Inc. (2013), PEM electrolyser incorporating
an advanced low-cost membrane, 2013 Hydrogen
Program Annual Merit Review Meeting, Giner Inc.
Governors Interagency Working Group on Zeroemission Vehicles (2013), ZEV Action Plan - A
Roadmap Toward 1.5 Million Zero-emission Vehicles
on California Roadways by 2025, Office of Governor
Edmund G. Brown Jr.
Hydrogen and Fuel Cell Strategy Council (2014),
Strategic Roadmap for Hydrogen Fuel Cells, Ministry
of Economy, Trade and Industry, Tokyo.
Hydrogen Implementing Agreement Task 25 (2009),
Alkaline Electrolysis, Hydrogen Implementing
Agreement.
HySUT (2014), Fuel Cell Vehicle Demonstration and
Hydrogen Infrastructure Project in Japan, HySUT.
HyUnder (2013), Assessment of the Potential, the
Actors and Relevant Business Cases for Large Scale and
Seasonal Storage of Renewable Electricity by Hydrogen
Underground Storage in Europe - Benchmarking of
Selected Storage Options, HyUnder.
IEA (International Energy Agency) (2015), Energy
Technology Perspectives, OECD/IEA, Paris.
IEA (2014a), Energy Technology Perspectives, OECD/
IEA, Paris.
IEA (2014b), Technology Roadmap: Energy Storage,
OECD/IEA, Paris.
IEA (2014c), The Power of Transformation, OECD/IEA,
Paris.
References
73
IEA (2013), Technology Roadmap for Carbon Capture

and Storage, OECD/IEA, Paris.
IEA (2012), Energy Technology Perspectives, OECD/
IEA, Paris.
IEA AFC IA (2014), IEA AFC IA Annex Meeting 25,
IEA Advanced Fuel Cells Implementing Agreement,
Trento.
IEA AFC IA (2015), International Status of Molten
Carbonate Fuel Cells Technology, IEA Advanced Fuel
Cells Implementing Agreement Annex 23 - MCFC,
Rome.
IFP Energies Nouvelles (2008), Raffinage 2030, IFP
Energies Nouvelles, Rueil-Malmaison.
Iiyama, A. et al. (2014), FCEV Development at
Nissan, ECS Transactions, Vol. 3, pp. 11-17.
IKA RWTH Aachen (n.d.), On-site Hydrogen
Generators from Hydrocarbons, www.ika.rwthaachen.de/r2h/index.php/On-site_Hydrogen_
Generators_from_Hydrocarbons (accessed
30March 2015).
Inoue, K. (2012), Feasibility Study of CO2 Free
Hydrogen Energy Supply Chain Utilizing Australian
Brown Coal Linked with CCS, submission to 12th
Cryogenics Conference, September 2012, Dresden.
Jorgensen, C. E. (2008), Production price of
hydrogen from grid connected electrolysis in
a power market with high wind penetration,
International Journal of Hydrogen Energy, pp.53355344.
JRC (Joint Research Centre) (2014), Final Report
Summary and Outcomes of the Workshop Putting
Science into Standards: Power-to-Hydrogen and
HCNG, Joint Research Centre, Petten.
JRC (2013), JRC Technical Reports Well-to-tank
Report Version 4.0 JEC Well-to-Wheels Analysis, Joint
Research Centre, Publication Office of the European
Union, Luxembourg.
Kopp, A. (2013), Power-to-gas and hydrogen within
E.on, IEA Hydrogen Roadmap Europe Workshop, E.on
Innovation Center Energy Storage, Paris.
Linde (n.d.), Hydrogen, www.linde-engineering.
com/internet.global.lindeengineering.global/en/
images/H2_1_1_e_12_150dpi19_4258.pdf, accessed
30March 2015.
74
Mansilla, C. E. (2013), Economic Competitiveness

of Off-peak Hydrogen Production Today - A
European Comparison, Energy, pp. 996-1001.
McKinsey and Co. (2012), Urban Buses: Alternative
Powertrains for Europe, McKinsey and Co.
McKinsey and Co. (2011), A Portfolio of Power-Trains
for Europe: a Fact-based Analysis, The Role of Battery
Electric Vehicles, Plug-in Hybrids and Fuel Cell Electric
Vehicles, McKinsey and Co.
National Research Council (2013), Transition to
Alternative Vehicles and Fuels, National Research
Council of the National Academies, Washington,
D.C.
NETL (2013), Carbon Dioxide Transport and Storage
Costs in NETL Studies, National Research Council.
NETL (National Energy Technology Laboratory)
(2010), Production of High Purity Hydrogen from
Domestic Coal: Assessing the Techno-Economic
Impact of Emerging Technologies, National Energy
Technology Laboratory.
NREL (National Renewable Energy Laboratory)
(2014), Hydrogen Station Compression, Storage and
Dispensing - Technical Status and Costs, National
Renewable Energy Laboratory, Golden.
NREL (2012a), National Fuel Cell Electric Vehicle
Learning Demonstration Final Report, National
Renewable Energy Laboratory, Golden.
NREL (2012b), Renewable Electricity Futures Study
- Exploration of High Renewable Electricity Futures
Volume 2, National Renewable Energy Laboratory,
Golden.
NREL (2010), Molten Carbonate and Phosphoric Acid
Stationary Fuel Cells: Overview and Gap Analysis,
National Renewable Energy Laboratory, Golden.
NREL (2009a), Scenario development and analysis
of hydrogen as a large-scale energy storage
medium, RMEL Meeting, National Renewable
Energy Laboratory, Denver.
NREL (2009b), Scenario Development and Analysis of
Hydrogen as a Large-Scale Energy Storage Medium,
National Renewable Energy Laboratory, Golden.
NEDO (New Energy and Industrial Technology
Development Organization) (2014), Hydrogen
Roadmap for Japan, New Energy and Industrial
Technology Development Organization.
Nexant (2007), Liquefaction and pipeline costs,

Hydrogen Delivery Analysis Meeting, 8-9May,
Nexant, Columbia, MD.
Ogden, J. et al. (2014), NextSTEPS White Paper: the
Hydrogen Transition, Institute of Transportation
Studies, University of California, Davis, CA.
OICA (Organisation Internationale des
Constructeurs dAutomobiles)(2014), Registration
or Sales of New Vehicles (12March), Organisation
Internationale des Constructeurs dAutomobiles.
Pacific Northwest National Laboratory (2015),
Hydrogen Pipelines Database, http://hydrogen.pnl.
gov/sites/default/files/data/Hydrogen%20Pipelines.
xls (accessed 8April), US Department of Energy.
USDOE (2014b), Hydrogen and Fuel Cells Program

Record, US Department of Energy, Washington, D.C.
USDOE (2014c), Hydrogen and Fuel Cells Program
Record - Industry Deployed Fuel Cell Powered Lift
Trucks, US Department of Energy, Washington, D.C.
USDOE (2014d), DOE Fuel Cell Technologies Office
Record Fuel Cell system Costs, US Department of
Energy, Washington, D.C.
USDOE (2013), Fuel Cell Technology Office Record
On-board Type IV Compressed Hydrogen Storage
Systems Current Performance and Cost, US
Department of Energy, Washington, D.C.
USDOE (2012), Fuel Cell Technologies Program
Record, US Department of Energy, Washington, D.C.
Pacific Northwest National Laboratory (2013),

Business Case for a Micro Combined Heat and Power
Fuel Cell System in Commercial Applications, US
Department of Energy.
USDOE (2011), Energy Hydrogen and Fuel

Cell Program Plan, US Department of Energy,
Washington, D.C.
Rabiei, Z. (2012), Hydrogen management in

refineries, Petroleum & Coal, pp. 357-368.
USDOE (2010a), Hydrogen Program 2010 Annual

Progress Report - Innovative Hydrogen Liquefaction
Cycle, US Department of Energy, Washington, D.C.
Saur, G. (2008), Wind-To-Hydrogen Project:

Electrolyzer Capital Cost Study, National Renewable
Energy Laboratory, Golden.
Schaber, K., F. Steinke and T. Hamacher (2013),
Managing temporary oversupply from renewables
efficiently: electricity storage versus energy sector
coupling in Germany, International Energy
Workshop, Paris.
USDOE (2010b), DOE Hydrogen Program 2010

Annual Progress Report - Technology Validation
Sub-Program Overview, US Department of Energy,
Washington, D.C.
Weeda, M. et al. (2014), Towards a Comprehensive
Hydrogen Infrastructure for Fuel Cell Electric Cars
in View of EU GHG Reduction Targets, Hydrogen
Infrastructure for Transport (HIT).
Stolzenburg, K. et al. (2014), Integration von

Wind-Wasserstoff-Systemen in das Energiesystem
Abschlussbericht, PLANET Planungsgruppe Energie
und Technik GbR.
Working Party on Passive Safety (2014), Proposal for

a New Regulation on Hydrogen and Fuel Cell Vehicles,
UNECE Economic and Social Council, Geneva.
Suresh, B. et al. (2013), Chemical Economics

Handbook, IHS Chemical.
World Bank (2014), Water and energy efficiency,

Green Bond: Sixth Annual Investor Update 2014,
Toyota (2015a), Toyota Mirai Performance, http://

toyota.jp/mirai/performance/, accessed 5May.
The World Bank Treasury, Washington, D.C.,

http://treasury.worldbank.org/cmd/pdf/
WorldBankGreenBondNewsletter.pdf.
Toyota (2015b), 2016 Toyota Mirai Fuel

Cell Sedan Product Information, http://
toyotanews.pressroom.toyota.com/releases/
toyota+mirai+fcv+future+nov17.htm, accessed
7April.
World Steel Association (2014), Steel Statistical

Yearbook 2014, World Steel Committee on Economic
Studies, Brussels.
USDOE (United States Department of Energy)

(2014a), 2013 Fuel Cell Technologies Market Report,
US Department of Energy, Washington, D.C.
Yang, C. and J. Ogden (2007), Determining the

lowest-cost hydrogen delivery mode, International
Journal of Hydrogen Energy, pp. 268-286.
References
75
nline
bookshop
www.iea.org/books
PDF versions
at 20% discount

9 rue de la Fdration
75739 Paris Cedex 15, France
Tel: +33 (0)1 40 57 66 90
E-mail: books@iea.org
Also from the IEA

Energy Technology
Perspectives 2015
Action
to Accelerate Climate
Mobilising Innovation
Energy Technology Perspectives is the International Energy Agencys

most ambitious project on new developments in energy
technology: its analysis and scenarios provide the benchmarks used
in Tracking Clean Energy Progress. Energy Technology Perspectives
2015 examines innovation in the energy technology sector and
seeks to increase confidence in the feasibility of achieving climate
change mitigation targets through effective RDD&D. ETP 2015
identifies strategies to advance innovation in areas like variable
renewables, CCS and energy-intensive industrial sectors. The report
also shows how emerging economies, and China in particular, can
foster a low-carbon transition through innovation in energy
technologies and policy. www.iea.org/etp2015
New releases
There are no quick fixes to long-term energy challenges. To find solutions, governments
and industry benefit from sharing resources and results. For this reason the IEA supports
multilateral experts groups on a wide range of technologies. Recent achievements of
these groups are synthesised in the biennial Energy Technology Initiatives. The 2015
edition marks the 40th anniversary. Forthcoming 4th quarter 2015.
www.iea.org/techinitiatives
The IEA Policies and Measures Databases contain information on policies and
measures to reduce greenhouse gas emissions, improve energy efficiency and support
renewable energy. This online service complements the policy analysis carried out by
the IEA and covers all IEA member countries as well as many non-member countries.
Delegates from IEA member countries are given the opportunity to review the
information twice a year. www.iea.org/policiesandmeasures
Capturing the Multiple Benefits of Energy Efficiency goes beyond the traditional analysis
of energy demand reduction and climate change mitigation. With its potential to
enhance economic prosperity, social development, sustainability and energy security,
the multiple benefits perspective can reposition energy efficiency in the policy
mainstream. www.iea.org/multiplebenefits
IEA Technology Roadmaps help set the agenda for development and deployment of
key low-carbon energy technologies. These free publications identify priority actions
needed to support technological advances and uptake by society in alignment with
the ETP 2DS vision. As of May 2015, 20 global roadmaps have been published, covering
topics including buildings, CCS, electric vehicles, nuclear, solar and storage. Further
global and regional roadmaps, as well as How2Guides, are in production. More details
at: www.iea.org/roadmaps
For more information, please visit www.iea.org
The paper used for this document has received certification from the Programme for the Endorsement of Forest Certification (PEFC) for being
produced respecting PEFCs ecological, social and ethical standards. PEFC is an international non-profit, non-governmental organization
dedicated to promoting Sustainable Forest Management (SFM) through independent third-party certification.
This document and any map included herein are without prejudice to the status of or sovereignty over any territory, to the delimitation of
international frontiers and boundaries and to the name of any territory, city or area.
IEA Publications, 9 rue de la Fdration, 75739 PARIS CEDEX 15

TYPESET BY IEA AND PRINTED IN FRANCE BY CORLET, JUNE 2015
Front cover photo (large): Flickr - Maurizio Pesce
Front cover photo (small): Flickr - Brewbooks
Back cover photo (large): Flickr - Lars Plougmann
Back cover photo (small): GraphicObsession
Technology Roadmap Hydrogen
2015
2016
2020
2025
2030
International Energy Agency IEA

9 rue de la Fdration, 75015 Paris, France
Tel: +33 (0)1 40 57 65 00
Fax: +33 (0)1 40 57 65 09
Email: info@iea.org, Web: www.iea.org

International Energy Agency Technology Roadmap H2 FuelCells

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

International Energy Agency Technology Roadmap H2 FuelCells

Загружено:

Авторское право:

Доступные форматы

2050

Hydrogen and Fuel Cells

INTERNATIONAL ENERGY AGENCY

IEA member countries:

The European Commission

Although the GHG mitigation potential of hydrogen

Maria van der Hoeven

Cross-cutting opportunities offered by hydrogen and fuel cells

Energy storage and utilisation in transport, industry and buildings

Key actions in the next ten years

Cross-cutting opportunities offered by hydrogen and fuel cells

Energy storage and utilisation in transport, industry and buildings

Rationale for hydrogen and fuel cell technologies

Purpose, process and structure of the roadmap

Technology status today

Hydrogen for VRE integration

Fuel cell technology in buildings

Other niche applications based on fuel cell technologies

Key hydrogen generation technologies

Key hydrogen conversion and storage technologies

Vision for deployment to 2050

Synergies between energy sectors

Hydrogen technology development: Actions and milestones

Data assessment and model development

Policy, regulatory framework and finance: Actions and milestones

Hydrogen in stationary applications

The role of codes and standards

Social acceptance and safety

Technology Roadmap Hydrogen and Fuel Cells

Conclusion: Near-term actions for stakeholders

Abbreviations, acronyms and units of measurement

Figure 5. Electricity storage applications and technologies

Figure 6. Current conversion efficiencies of various hydrogen-based VRE integration pathways

Figure 7. Limitations on the blend share of hydrogen by application

Figure 10. Production volumes of fuel cells according to application

Figure 28. Electricity price arbitrage and hydrogen generation costs

Table 2. Current performance of hydrogen systems in the transport sector

Table 3. Existing FCEV fleet and targets announced by hydrogen initiatives

Table 6. Current performance of hydrogen systems for large-scale energy storage

Table 8. Current performance of fuel cell systems in the buildings sector

Table 9. Current performance of key hydrogen generation technologies

Table 13. Power-to-power and power-to-gas systems included in the analysis

Box 1. Risks associated with investment in hydrogen refuelling stations

Box 2. Carbon footprint of hydrogen in transport

Box 3. Power-to-gas in Europe: storage potential and limitations

Box 4. The Japanese Ene-Farm experience

Box 5. ETP scenarios and the hydrogen roadmap variant

Box 6. The economics of renewable hydrogen

Box 7. Spotlight on hydrogen generation

Box 8. Vehicle costs, fuel cost and TCD

Box 9. Electrolysers in the control power market

Technology Roadmap Hydrogen and Fuel Cells

Finally, the IEA would like to thank the numerous

of hydrogen can be stored over long periods of

1. Unless otherwise stated, all monetary values are in 2013 USD.

Energy storage and

Technology Roadmap Hydrogen and Fuel Cells

Key actions in the next ten years

Energy storage and

zz E ncourage fuel efficiency and low greenhouse gas

zz E ngage international stakeholders from relevant

zz Increase the number of hydrogen-based energy

zz E stablish regulatory frameworks that remove

zz Improve understanding of regionally specific

1. Unless otherwise stated, all monetary values are in 2013 USD.

zz E ncourage fuel efficiency and low greenhouse gas

zz E ngage international stakeholders from relevant

zz Increase the number of hydrogen-based energy

zz E stablish regulatory frameworks that remove

zz Improve understanding of regionally specific

zz Increase data on resource availability and costs

2. Co-generation refers to the combined production of heat

zz Identify applications where using hydrogen can

3. In 2013 around 63millionpassenger light-duty vehicles